{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.2000842094421387, "learning_rate": 2.25e-07, "loss": 1.4409, "step": 10 }, { "grad_norm": 1.2061761617660522, "learning_rate": 4.75e-07, "loss": 1.4262, "step": 20 }, { "grad_norm": 1.2710939645767212, "learning_rate": 7.25e-07, "loss": 1.4332, "step": 30 }, { "grad_norm": 1.0610424280166626, "learning_rate": 9.75e-07, "loss": 1.4414, "step": 40 }, { "grad_norm": 1.1537632942199707, "learning_rate": 1.2250000000000001e-06, "loss": 1.4225, "step": 50 }, { "grad_norm": 1.7849572896957397, "learning_rate": 1.475e-06, "loss": 1.3705, "step": 60 }, { "grad_norm": 1.7827318906784058, "learning_rate": 1.7250000000000002e-06, "loss": 1.2422, "step": 70 }, { "grad_norm": 2.2491817474365234, "learning_rate": 1.975e-06, "loss": 1.2061, "step": 80 }, { "grad_norm": 1.4837195873260498, "learning_rate": 2.225e-06, "loss": 1.2341, "step": 90 }, { "grad_norm": 2.0516672134399414, "learning_rate": 2.4750000000000004e-06, "loss": 1.174, "step": 100 }, { "grad_norm": 1.5020558834075928, "learning_rate": 2.725e-06, "loss": 1.147, "step": 110 }, { "grad_norm": 1.5903315544128418, "learning_rate": 2.975e-06, "loss": 1.1552, "step": 120 }, { "grad_norm": 1.5490655899047852, "learning_rate": 3.225e-06, "loss": 1.1246, "step": 130 }, { "grad_norm": 1.6885154247283936, "learning_rate": 3.4750000000000006e-06, "loss": 1.1237, "step": 140 }, { "grad_norm": 1.6795251369476318, "learning_rate": 3.725e-06, "loss": 1.0759, "step": 150 }, { "grad_norm": 1.9751113653182983, "learning_rate": 3.975e-06, "loss": 1.075, "step": 160 }, { "grad_norm": 1.5594308376312256, "learning_rate": 4.225e-06, "loss": 1.0679, "step": 170 }, { "grad_norm": 1.4654279947280884, "learning_rate": 4.475e-06, "loss": 1.0637, "step": 180 }, { "grad_norm": 1.350258708000183, "learning_rate": 4.7250000000000005e-06, "loss": 1.0797, "step": 190 }, { "grad_norm": 1.912964940071106, "learning_rate": 4.975000000000001e-06, "loss": 1.064, "step": 200 }, { "grad_norm": 1.3728253841400146, "learning_rate": 5.225e-06, "loss": 1.0424, "step": 210 }, { "grad_norm": 1.255905270576477, "learning_rate": 5.475e-06, "loss": 1.0678, "step": 220 }, { "grad_norm": 1.4538661241531372, "learning_rate": 5.725e-06, "loss": 1.0515, "step": 230 }, { "grad_norm": 1.5353754758834839, "learning_rate": 5.975e-06, "loss": 1.0538, "step": 240 }, { "grad_norm": 1.2111200094223022, "learning_rate": 6.2250000000000005e-06, "loss": 1.047, "step": 250 }, { "grad_norm": 1.1821489334106445, "learning_rate": 6.475000000000001e-06, "loss": 1.0256, "step": 260 }, { "grad_norm": 1.5669541358947754, "learning_rate": 6.725000000000001e-06, "loss": 1.0489, "step": 270 }, { "grad_norm": 1.3441468477249146, "learning_rate": 6.975000000000001e-06, "loss": 1.0396, "step": 280 }, { "grad_norm": 1.6470670700073242, "learning_rate": 7.2249999999999994e-06, "loss": 1.038, "step": 290 }, { "grad_norm": 1.1853636503219604, "learning_rate": 7.4750000000000004e-06, "loss": 1.0293, "step": 300 }, { "grad_norm": 1.2297625541687012, "learning_rate": 7.725e-06, "loss": 1.0502, "step": 310 }, { "grad_norm": 1.9184446334838867, "learning_rate": 7.975e-06, "loss": 1.0307, "step": 320 }, { "grad_norm": 1.3753594160079956, "learning_rate": 8.225e-06, "loss": 1.0224, "step": 330 }, { "grad_norm": 2.8201067447662354, "learning_rate": 8.475000000000001e-06, "loss": 1.0174, "step": 340 }, { "grad_norm": 2.005849599838257, "learning_rate": 8.725e-06, "loss": 1.0016, "step": 350 }, { "grad_norm": 1.7425307035446167, "learning_rate": 8.975e-06, "loss": 0.9678, "step": 360 }, { "grad_norm": 2.076201915740967, "learning_rate": 9.225e-06, "loss": 0.9688, "step": 370 }, { "grad_norm": 2.482668399810791, "learning_rate": 9.475e-06, "loss": 0.9271, "step": 380 }, { "grad_norm": 2.0649943351745605, "learning_rate": 9.725000000000001e-06, "loss": 0.8976, "step": 390 }, { "grad_norm": 2.511237621307373, "learning_rate": 9.975e-06, "loss": 0.8794, "step": 400 }, { "grad_norm": 2.4416234493255615, "learning_rate": 1.0225e-05, "loss": 0.8405, "step": 410 }, { "grad_norm": 2.376070737838745, "learning_rate": 1.0475e-05, "loss": 0.8174, "step": 420 }, { "grad_norm": 2.509124994277954, "learning_rate": 1.0725e-05, "loss": 0.7827, "step": 430 }, { "grad_norm": 2.3622241020202637, "learning_rate": 1.0975e-05, "loss": 0.775, "step": 440 }, { "grad_norm": 2.514601945877075, "learning_rate": 1.1225e-05, "loss": 0.7171, "step": 450 }, { "grad_norm": 2.5513980388641357, "learning_rate": 1.1475000000000001e-05, "loss": 0.7105, "step": 460 }, { "grad_norm": 2.3124167919158936, "learning_rate": 1.1725e-05, "loss": 0.6968, "step": 470 }, { "grad_norm": 2.765599250793457, "learning_rate": 1.1975e-05, "loss": 0.6842, "step": 480 }, { "grad_norm": 2.74357271194458, "learning_rate": 1.2225e-05, "loss": 0.6484, "step": 490 }, { "grad_norm": 3.2130050659179688, "learning_rate": 1.2475e-05, "loss": 0.6152, "step": 500 }, { "grad_norm": 2.6422650814056396, "learning_rate": 1.2725000000000001e-05, "loss": 0.6021, "step": 510 }, { "grad_norm": 2.814192056655884, "learning_rate": 1.2975e-05, "loss": 0.5917, "step": 520 }, { "grad_norm": 3.0668115615844727, "learning_rate": 1.3225000000000001e-05, "loss": 0.5827, "step": 530 }, { "grad_norm": 2.3118438720703125, "learning_rate": 1.3475000000000002e-05, "loss": 0.5022, "step": 540 }, { "grad_norm": 2.9654698371887207, "learning_rate": 1.3725000000000002e-05, "loss": 0.5406, "step": 550 }, { "grad_norm": 3.240199565887451, "learning_rate": 1.3975000000000003e-05, "loss": 0.5222, "step": 560 }, { "grad_norm": 2.865651845932007, "learning_rate": 1.4225e-05, "loss": 0.4664, "step": 570 }, { "grad_norm": 2.94984769821167, "learning_rate": 1.4475e-05, "loss": 0.452, "step": 580 }, { "grad_norm": 3.109316110610962, "learning_rate": 1.4725e-05, "loss": 0.4359, "step": 590 }, { "grad_norm": 2.3523616790771484, "learning_rate": 1.4975e-05, "loss": 0.4243, "step": 600 }, { "grad_norm": 2.1841745376586914, "learning_rate": 1.5225e-05, "loss": 0.4162, "step": 610 }, { "grad_norm": 3.533320426940918, "learning_rate": 1.5475e-05, "loss": 0.3957, "step": 620 }, { "grad_norm": 2.3309402465820312, "learning_rate": 1.5725e-05, "loss": 0.3976, "step": 630 }, { "grad_norm": 3.0528767108917236, "learning_rate": 1.5975000000000002e-05, "loss": 0.3622, "step": 640 }, { "grad_norm": 3.0812432765960693, "learning_rate": 1.6225e-05, "loss": 0.3467, "step": 650 }, { "grad_norm": 3.310957670211792, "learning_rate": 1.6475e-05, "loss": 0.3539, "step": 660 }, { "grad_norm": 2.1877732276916504, "learning_rate": 1.6725000000000003e-05, "loss": 0.3188, "step": 670 }, { "grad_norm": 2.67159366607666, "learning_rate": 1.6975000000000003e-05, "loss": 0.3122, "step": 680 }, { "grad_norm": 2.6558501720428467, "learning_rate": 1.7225e-05, "loss": 0.299, "step": 690 }, { "grad_norm": 3.431931972503662, "learning_rate": 1.7475e-05, "loss": 0.2983, "step": 700 }, { "grad_norm": 3.454773187637329, "learning_rate": 1.7725e-05, "loss": 0.2675, "step": 710 }, { "grad_norm": 3.311849355697632, "learning_rate": 1.7975e-05, "loss": 0.2709, "step": 720 }, { "grad_norm": 2.6970713138580322, "learning_rate": 1.8225e-05, "loss": 0.2442, "step": 730 }, { "grad_norm": 2.2823448181152344, "learning_rate": 1.8475000000000002e-05, "loss": 0.214, "step": 740 }, { "grad_norm": 2.6653871536254883, "learning_rate": 1.8725e-05, "loss": 0.2339, "step": 750 }, { "grad_norm": 2.454246759414673, "learning_rate": 1.8975e-05, "loss": 0.2214, "step": 760 }, { "grad_norm": 2.5415496826171875, "learning_rate": 1.9225e-05, "loss": 0.2218, "step": 770 }, { "grad_norm": 2.5333876609802246, "learning_rate": 1.9475000000000002e-05, "loss": 0.2275, "step": 780 }, { "grad_norm": 2.040771722793579, "learning_rate": 1.9725000000000002e-05, "loss": 0.187, "step": 790 }, { "grad_norm": 2.6439924240112305, "learning_rate": 1.9975e-05, "loss": 0.163, "step": 800 }, { "grad_norm": 3.2696244716644287, "learning_rate": 2.0225000000000004e-05, "loss": 0.176, "step": 810 }, { "grad_norm": 2.0922281742095947, "learning_rate": 2.0475e-05, "loss": 0.1947, "step": 820 }, { "grad_norm": 2.6732046604156494, "learning_rate": 2.0725e-05, "loss": 0.2149, "step": 830 }, { "grad_norm": 2.048996686935425, "learning_rate": 2.0975e-05, "loss": 0.1967, "step": 840 }, { "grad_norm": 2.172708511352539, "learning_rate": 2.1225e-05, "loss": 0.2007, "step": 850 }, { "grad_norm": 2.7855939865112305, "learning_rate": 2.1475e-05, "loss": 0.169, "step": 860 }, { "grad_norm": 2.4105703830718994, "learning_rate": 2.1725e-05, "loss": 0.2001, "step": 870 }, { "grad_norm": 2.855689764022827, "learning_rate": 2.1975000000000002e-05, "loss": 0.176, "step": 880 }, { "grad_norm": 2.00956392288208, "learning_rate": 2.2225e-05, "loss": 0.1783, "step": 890 }, { "grad_norm": 2.3201637268066406, "learning_rate": 2.2475e-05, "loss": 0.1772, "step": 900 }, { "grad_norm": 1.725034236907959, "learning_rate": 2.2725000000000003e-05, "loss": 0.1568, "step": 910 }, { "grad_norm": 2.06207537651062, "learning_rate": 2.2975000000000003e-05, "loss": 0.1529, "step": 920 }, { "grad_norm": 2.5735390186309814, "learning_rate": 2.3225000000000002e-05, "loss": 0.1559, "step": 930 }, { "grad_norm": 2.614659547805786, "learning_rate": 2.3475e-05, "loss": 0.1694, "step": 940 }, { "grad_norm": 2.7896132469177246, "learning_rate": 2.3725e-05, "loss": 0.1276, "step": 950 }, { "grad_norm": 2.2075672149658203, "learning_rate": 2.3975e-05, "loss": 0.1549, "step": 960 }, { "grad_norm": 2.2386724948883057, "learning_rate": 2.4225e-05, "loss": 0.1764, "step": 970 }, { "grad_norm": 2.1529998779296875, "learning_rate": 2.4475000000000002e-05, "loss": 0.1446, "step": 980 }, { "grad_norm": 2.586261510848999, "learning_rate": 2.4725e-05, "loss": 0.1532, "step": 990 }, { "grad_norm": 2.0143837928771973, "learning_rate": 2.4975e-05, "loss": 0.1429, "step": 1000 }, { "grad_norm": 1.9130194187164307, "learning_rate": 2.5225e-05, "loss": 0.1495, "step": 1010 }, { "grad_norm": 2.608229875564575, "learning_rate": 2.5475e-05, "loss": 0.1666, "step": 1020 }, { "grad_norm": 2.0579605102539062, "learning_rate": 2.5725e-05, "loss": 0.1304, "step": 1030 }, { "grad_norm": 2.075793743133545, "learning_rate": 2.5974999999999998e-05, "loss": 0.1312, "step": 1040 }, { "grad_norm": 2.3542909622192383, "learning_rate": 2.6225e-05, "loss": 0.1475, "step": 1050 }, { "grad_norm": 2.9262290000915527, "learning_rate": 2.6475e-05, "loss": 0.1608, "step": 1060 }, { "grad_norm": 2.376093626022339, "learning_rate": 2.6725e-05, "loss": 0.1721, "step": 1070 }, { "grad_norm": 2.649322986602783, "learning_rate": 2.6975000000000002e-05, "loss": 0.1839, "step": 1080 }, { "grad_norm": 1.594651699066162, "learning_rate": 2.7225e-05, "loss": 0.1608, "step": 1090 }, { "grad_norm": 1.8970260620117188, "learning_rate": 2.7475e-05, "loss": 0.1305, "step": 1100 }, { "grad_norm": 2.326268196105957, "learning_rate": 2.7725e-05, "loss": 0.1795, "step": 1110 }, { "grad_norm": 1.8001340627670288, "learning_rate": 2.7975000000000002e-05, "loss": 0.1131, "step": 1120 }, { "grad_norm": 1.7213122844696045, "learning_rate": 2.8225e-05, "loss": 0.1534, "step": 1130 }, { "grad_norm": 2.6074047088623047, "learning_rate": 2.8475e-05, "loss": 0.1821, "step": 1140 }, { "grad_norm": 2.129626512527466, "learning_rate": 2.8725e-05, "loss": 0.181, "step": 1150 }, { "grad_norm": 1.3883869647979736, "learning_rate": 2.8975000000000003e-05, "loss": 0.1494, "step": 1160 }, { "grad_norm": 2.9133119583129883, "learning_rate": 2.9225000000000002e-05, "loss": 0.15, "step": 1170 }, { "grad_norm": 1.7240726947784424, "learning_rate": 2.9475e-05, "loss": 0.1242, "step": 1180 }, { "grad_norm": 1.7783383131027222, "learning_rate": 2.9725000000000004e-05, "loss": 0.1483, "step": 1190 }, { "grad_norm": 3.26481032371521, "learning_rate": 2.9975000000000004e-05, "loss": 0.1485, "step": 1200 }, { "grad_norm": 3.362461566925049, "learning_rate": 3.0225000000000003e-05, "loss": 0.1708, "step": 1210 }, { "grad_norm": 1.8677726984024048, "learning_rate": 3.0475000000000002e-05, "loss": 0.1442, "step": 1220 }, { "grad_norm": 1.7161425352096558, "learning_rate": 3.0725e-05, "loss": 0.1441, "step": 1230 }, { "grad_norm": 1.6996288299560547, "learning_rate": 3.0975e-05, "loss": 0.1544, "step": 1240 }, { "grad_norm": 1.8923254013061523, "learning_rate": 3.122500000000001e-05, "loss": 0.1304, "step": 1250 }, { "grad_norm": 1.9257216453552246, "learning_rate": 3.1475e-05, "loss": 0.159, "step": 1260 }, { "grad_norm": 2.138754367828369, "learning_rate": 3.1725e-05, "loss": 0.1278, "step": 1270 }, { "grad_norm": 1.5533071756362915, "learning_rate": 3.1975e-05, "loss": 0.1569, "step": 1280 }, { "grad_norm": 1.1481308937072754, "learning_rate": 3.2225e-05, "loss": 0.1376, "step": 1290 }, { "grad_norm": 1.8471814393997192, "learning_rate": 3.2474999999999997e-05, "loss": 0.1373, "step": 1300 }, { "grad_norm": 1.7611762285232544, "learning_rate": 3.2725e-05, "loss": 0.1314, "step": 1310 }, { "grad_norm": 2.5457756519317627, "learning_rate": 3.2975e-05, "loss": 0.1341, "step": 1320 }, { "grad_norm": 1.808872103691101, "learning_rate": 3.3225e-05, "loss": 0.1191, "step": 1330 }, { "grad_norm": 1.6629351377487183, "learning_rate": 3.3475e-05, "loss": 0.1268, "step": 1340 }, { "grad_norm": 1.7607173919677734, "learning_rate": 3.3725e-05, "loss": 0.155, "step": 1350 }, { "grad_norm": 2.474785804748535, "learning_rate": 3.3975e-05, "loss": 0.1349, "step": 1360 }, { "grad_norm": 1.8439346551895142, "learning_rate": 3.4225e-05, "loss": 0.1249, "step": 1370 }, { "grad_norm": 1.4883534908294678, "learning_rate": 3.4475000000000005e-05, "loss": 0.116, "step": 1380 }, { "grad_norm": 1.7489567995071411, "learning_rate": 3.4725000000000004e-05, "loss": 0.1291, "step": 1390 }, { "grad_norm": 1.573334813117981, "learning_rate": 3.4975e-05, "loss": 0.1201, "step": 1400 }, { "grad_norm": 1.095489501953125, "learning_rate": 3.5225e-05, "loss": 0.1301, "step": 1410 }, { "grad_norm": 1.7740620374679565, "learning_rate": 3.5475e-05, "loss": 0.1265, "step": 1420 }, { "grad_norm": 1.3462629318237305, "learning_rate": 3.5725e-05, "loss": 0.0986, "step": 1430 }, { "grad_norm": 1.8774449825286865, "learning_rate": 3.5975e-05, "loss": 0.1427, "step": 1440 }, { "grad_norm": 1.9436064958572388, "learning_rate": 3.6225000000000006e-05, "loss": 0.1406, "step": 1450 }, { "grad_norm": 1.454484462738037, "learning_rate": 3.6475000000000006e-05, "loss": 0.1459, "step": 1460 }, { "grad_norm": 1.672963261604309, "learning_rate": 3.6725000000000005e-05, "loss": 0.1432, "step": 1470 }, { "grad_norm": 1.5753543376922607, "learning_rate": 3.6975000000000004e-05, "loss": 0.125, "step": 1480 }, { "grad_norm": 1.5648303031921387, "learning_rate": 3.7225000000000004e-05, "loss": 0.1268, "step": 1490 }, { "grad_norm": 1.1841092109680176, "learning_rate": 3.7475e-05, "loss": 0.1223, "step": 1500 }, { "grad_norm": 1.4961618185043335, "learning_rate": 3.7725e-05, "loss": 0.1092, "step": 1510 }, { "grad_norm": 2.8824992179870605, "learning_rate": 3.7975e-05, "loss": 0.1375, "step": 1520 }, { "grad_norm": 1.4380908012390137, "learning_rate": 3.8225e-05, "loss": 0.1382, "step": 1530 }, { "grad_norm": 1.3178907632827759, "learning_rate": 3.8475e-05, "loss": 0.1261, "step": 1540 }, { "grad_norm": 1.3017959594726562, "learning_rate": 3.8725e-05, "loss": 0.1176, "step": 1550 }, { "grad_norm": 1.4099841117858887, "learning_rate": 3.8975e-05, "loss": 0.1305, "step": 1560 }, { "grad_norm": 1.316420316696167, "learning_rate": 3.9225e-05, "loss": 0.1048, "step": 1570 }, { "grad_norm": 1.3568484783172607, "learning_rate": 3.9475000000000004e-05, "loss": 0.126, "step": 1580 }, { "grad_norm": 1.4111956357955933, "learning_rate": 3.9725e-05, "loss": 0.1285, "step": 1590 }, { "grad_norm": 1.211941123008728, "learning_rate": 3.9975e-05, "loss": 0.1104, "step": 1600 }, { "grad_norm": 1.2184138298034668, "learning_rate": 4.0225e-05, "loss": 0.106, "step": 1610 }, { "grad_norm": 1.233905553817749, "learning_rate": 4.0475e-05, "loss": 0.128, "step": 1620 }, { "grad_norm": 1.3379757404327393, "learning_rate": 4.0725e-05, "loss": 0.1536, "step": 1630 }, { "grad_norm": 1.2242212295532227, "learning_rate": 4.0975e-05, "loss": 0.132, "step": 1640 }, { "grad_norm": 1.5479122400283813, "learning_rate": 4.1225e-05, "loss": 0.1173, "step": 1650 }, { "grad_norm": 1.7782734632492065, "learning_rate": 4.1475000000000005e-05, "loss": 0.1327, "step": 1660 }, { "grad_norm": 1.404347538948059, "learning_rate": 4.1725000000000005e-05, "loss": 0.1139, "step": 1670 }, { "grad_norm": 1.4940382242202759, "learning_rate": 4.1975000000000004e-05, "loss": 0.1179, "step": 1680 }, { "grad_norm": 1.3694119453430176, "learning_rate": 4.2225e-05, "loss": 0.1196, "step": 1690 }, { "grad_norm": 0.9999154210090637, "learning_rate": 4.2475e-05, "loss": 0.1198, "step": 1700 }, { "grad_norm": 2.269373655319214, "learning_rate": 4.2725e-05, "loss": 0.1566, "step": 1710 }, { "grad_norm": 1.3772263526916504, "learning_rate": 4.2975e-05, "loss": 0.1221, "step": 1720 }, { "grad_norm": 1.2958993911743164, "learning_rate": 4.322500000000001e-05, "loss": 0.1172, "step": 1730 }, { "grad_norm": 1.3131449222564697, "learning_rate": 4.3475000000000006e-05, "loss": 0.118, "step": 1740 }, { "grad_norm": 1.5251185894012451, "learning_rate": 4.3725000000000006e-05, "loss": 0.1338, "step": 1750 }, { "grad_norm": 1.5831066370010376, "learning_rate": 4.3975e-05, "loss": 0.126, "step": 1760 }, { "grad_norm": 1.141121745109558, "learning_rate": 4.4225e-05, "loss": 0.1023, "step": 1770 }, { "grad_norm": 1.1481648683547974, "learning_rate": 4.4475e-05, "loss": 0.1303, "step": 1780 }, { "grad_norm": 1.6419438123703003, "learning_rate": 4.4725e-05, "loss": 0.1067, "step": 1790 }, { "grad_norm": 1.174017310142517, "learning_rate": 4.4975e-05, "loss": 0.1207, "step": 1800 }, { "grad_norm": 1.1847878694534302, "learning_rate": 4.5225e-05, "loss": 0.1116, "step": 1810 }, { "grad_norm": 1.4258240461349487, "learning_rate": 4.5475e-05, "loss": 0.1236, "step": 1820 }, { "grad_norm": 1.6910903453826904, "learning_rate": 4.5725e-05, "loss": 0.1237, "step": 1830 }, { "grad_norm": 1.3676865100860596, "learning_rate": 4.5975e-05, "loss": 0.1413, "step": 1840 }, { "grad_norm": 1.098799705505371, "learning_rate": 4.6225e-05, "loss": 0.1452, "step": 1850 }, { "grad_norm": 1.189947247505188, "learning_rate": 4.6475000000000005e-05, "loss": 0.1489, "step": 1860 }, { "grad_norm": 1.0974172353744507, "learning_rate": 4.6725000000000004e-05, "loss": 0.1341, "step": 1870 }, { "grad_norm": 1.4672290086746216, "learning_rate": 4.6975000000000003e-05, "loss": 0.1028, "step": 1880 }, { "grad_norm": 1.4914175271987915, "learning_rate": 4.7225e-05, "loss": 0.1319, "step": 1890 }, { "grad_norm": 0.946414589881897, "learning_rate": 4.7475e-05, "loss": 0.1106, "step": 1900 }, { "grad_norm": 1.512268304824829, "learning_rate": 4.7725e-05, "loss": 0.1452, "step": 1910 }, { "grad_norm": 1.243078351020813, "learning_rate": 4.7975e-05, "loss": 0.1222, "step": 1920 }, { "grad_norm": 0.9428762197494507, "learning_rate": 4.822500000000001e-05, "loss": 0.1084, "step": 1930 }, { "grad_norm": 1.5213661193847656, "learning_rate": 4.8475000000000006e-05, "loss": 0.1204, "step": 1940 }, { "grad_norm": 0.7905174493789673, "learning_rate": 4.8725000000000005e-05, "loss": 0.0971, "step": 1950 }, { "grad_norm": 1.7915894985198975, "learning_rate": 4.8975000000000005e-05, "loss": 0.1444, "step": 1960 }, { "grad_norm": 0.9406207203865051, "learning_rate": 4.9225000000000004e-05, "loss": 0.0922, "step": 1970 }, { "grad_norm": 1.6526038646697998, "learning_rate": 4.9475e-05, "loss": 0.1225, "step": 1980 }, { "grad_norm": 1.0247867107391357, "learning_rate": 4.9725e-05, "loss": 0.1044, "step": 1990 }, { "grad_norm": 1.2824645042419434, "learning_rate": 4.9975e-05, "loss": 0.119, "step": 2000 }, { "grad_norm": 1.1618645191192627, "learning_rate": 5.0225e-05, "loss": 0.136, "step": 2010 }, { "grad_norm": 1.2159960269927979, "learning_rate": 5.047500000000001e-05, "loss": 0.119, "step": 2020 }, { "grad_norm": 1.4296060800552368, "learning_rate": 5.0725e-05, "loss": 0.1053, "step": 2030 }, { "grad_norm": 0.9308567643165588, "learning_rate": 5.0975000000000006e-05, "loss": 0.1189, "step": 2040 }, { "grad_norm": 1.0867999792099, "learning_rate": 5.1225e-05, "loss": 0.1197, "step": 2050 }, { "grad_norm": 1.213399887084961, "learning_rate": 5.1475000000000004e-05, "loss": 0.1053, "step": 2060 }, { "grad_norm": 1.4811829328536987, "learning_rate": 5.1725000000000004e-05, "loss": 0.1152, "step": 2070 }, { "grad_norm": 1.3942718505859375, "learning_rate": 5.197500000000001e-05, "loss": 0.1135, "step": 2080 }, { "grad_norm": 1.1290009021759033, "learning_rate": 5.2225e-05, "loss": 0.1239, "step": 2090 }, { "grad_norm": 0.9465420246124268, "learning_rate": 5.247500000000001e-05, "loss": 0.1134, "step": 2100 }, { "grad_norm": 1.048182487487793, "learning_rate": 5.2725e-05, "loss": 0.1326, "step": 2110 }, { "grad_norm": 1.722839593887329, "learning_rate": 5.297500000000001e-05, "loss": 0.1146, "step": 2120 }, { "grad_norm": 0.8130356669425964, "learning_rate": 5.3225e-05, "loss": 0.1083, "step": 2130 }, { "grad_norm": 1.0925562381744385, "learning_rate": 5.3475e-05, "loss": 0.0978, "step": 2140 }, { "grad_norm": 1.2225325107574463, "learning_rate": 5.3725000000000005e-05, "loss": 0.1356, "step": 2150 }, { "grad_norm": 1.1041653156280518, "learning_rate": 5.3975e-05, "loss": 0.1199, "step": 2160 }, { "grad_norm": 0.7425931096076965, "learning_rate": 5.4225000000000003e-05, "loss": 0.1044, "step": 2170 }, { "grad_norm": 1.1402653455734253, "learning_rate": 5.4474999999999996e-05, "loss": 0.1251, "step": 2180 }, { "grad_norm": 1.0368064641952515, "learning_rate": 5.4725e-05, "loss": 0.1023, "step": 2190 }, { "grad_norm": 0.7456842064857483, "learning_rate": 5.4975e-05, "loss": 0.0945, "step": 2200 }, { "grad_norm": 0.8182862401008606, "learning_rate": 5.522500000000001e-05, "loss": 0.0937, "step": 2210 }, { "grad_norm": 1.086647868156433, "learning_rate": 5.5475e-05, "loss": 0.0932, "step": 2220 }, { "grad_norm": 1.047824740409851, "learning_rate": 5.5725000000000006e-05, "loss": 0.1099, "step": 2230 }, { "grad_norm": 1.0224575996398926, "learning_rate": 5.5975e-05, "loss": 0.103, "step": 2240 }, { "grad_norm": 0.8300219178199768, "learning_rate": 5.6225000000000005e-05, "loss": 0.1353, "step": 2250 }, { "grad_norm": 1.4796016216278076, "learning_rate": 5.6475e-05, "loss": 0.1008, "step": 2260 }, { "grad_norm": 0.8591707944869995, "learning_rate": 5.6725e-05, "loss": 0.1076, "step": 2270 }, { "grad_norm": 1.0015610456466675, "learning_rate": 5.6975e-05, "loss": 0.1056, "step": 2280 }, { "grad_norm": 1.3620871305465698, "learning_rate": 5.722500000000001e-05, "loss": 0.1537, "step": 2290 }, { "grad_norm": 1.2097547054290771, "learning_rate": 5.7475e-05, "loss": 0.1029, "step": 2300 }, { "grad_norm": 0.6787731647491455, "learning_rate": 5.772500000000001e-05, "loss": 0.0971, "step": 2310 }, { "grad_norm": 1.019882082939148, "learning_rate": 5.7975e-05, "loss": 0.0951, "step": 2320 }, { "grad_norm": 0.7826066017150879, "learning_rate": 5.8225000000000006e-05, "loss": 0.1093, "step": 2330 }, { "grad_norm": 0.7966407537460327, "learning_rate": 5.8475000000000005e-05, "loss": 0.1029, "step": 2340 }, { "grad_norm": 0.8432629704475403, "learning_rate": 5.8725000000000004e-05, "loss": 0.1133, "step": 2350 }, { "grad_norm": 1.1126402616500854, "learning_rate": 5.8975000000000004e-05, "loss": 0.1055, "step": 2360 }, { "grad_norm": 1.0446382761001587, "learning_rate": 5.922500000000001e-05, "loss": 0.118, "step": 2370 }, { "grad_norm": 1.0085049867630005, "learning_rate": 5.9475e-05, "loss": 0.1229, "step": 2380 }, { "grad_norm": 0.8901163339614868, "learning_rate": 5.9724999999999995e-05, "loss": 0.1175, "step": 2390 }, { "grad_norm": 1.040358304977417, "learning_rate": 5.9975e-05, "loss": 0.1054, "step": 2400 }, { "grad_norm": 0.8737033605575562, "learning_rate": 6.0225e-05, "loss": 0.1017, "step": 2410 }, { "grad_norm": 0.9130236506462097, "learning_rate": 6.0475000000000006e-05, "loss": 0.1417, "step": 2420 }, { "grad_norm": 1.5938835144042969, "learning_rate": 6.0725e-05, "loss": 0.124, "step": 2430 }, { "grad_norm": 1.2521331310272217, "learning_rate": 6.0975000000000005e-05, "loss": 0.1248, "step": 2440 }, { "grad_norm": 1.0673420429229736, "learning_rate": 6.1225e-05, "loss": 0.1183, "step": 2450 }, { "grad_norm": 1.0465227365493774, "learning_rate": 6.1475e-05, "loss": 0.103, "step": 2460 }, { "grad_norm": 1.0016556978225708, "learning_rate": 6.1725e-05, "loss": 0.1033, "step": 2470 }, { "grad_norm": 0.8277531266212463, "learning_rate": 6.1975e-05, "loss": 0.1175, "step": 2480 }, { "grad_norm": 0.9905164837837219, "learning_rate": 6.2225e-05, "loss": 0.1017, "step": 2490 }, { "grad_norm": 0.6855323314666748, "learning_rate": 6.2475e-05, "loss": 0.092, "step": 2500 }, { "grad_norm": 0.9623692631721497, "learning_rate": 6.2725e-05, "loss": 0.092, "step": 2510 }, { "grad_norm": 0.8004727363586426, "learning_rate": 6.297500000000001e-05, "loss": 0.1026, "step": 2520 }, { "grad_norm": 0.7947219610214233, "learning_rate": 6.3225e-05, "loss": 0.0883, "step": 2530 }, { "grad_norm": 0.6886498332023621, "learning_rate": 6.347500000000001e-05, "loss": 0.0919, "step": 2540 }, { "grad_norm": 1.1540932655334473, "learning_rate": 6.3725e-05, "loss": 0.114, "step": 2550 }, { "grad_norm": 0.6909359693527222, "learning_rate": 6.397500000000001e-05, "loss": 0.1188, "step": 2560 }, { "grad_norm": 0.9430059790611267, "learning_rate": 6.4225e-05, "loss": 0.1051, "step": 2570 }, { "grad_norm": 1.0031516551971436, "learning_rate": 6.447500000000001e-05, "loss": 0.1059, "step": 2580 }, { "grad_norm": 0.9815488457679749, "learning_rate": 6.4725e-05, "loss": 0.0978, "step": 2590 }, { "grad_norm": 1.1906161308288574, "learning_rate": 6.497500000000001e-05, "loss": 0.1181, "step": 2600 }, { "grad_norm": 0.7569696307182312, "learning_rate": 6.5225e-05, "loss": 0.1073, "step": 2610 }, { "grad_norm": 0.7455458641052246, "learning_rate": 6.5475e-05, "loss": 0.0929, "step": 2620 }, { "grad_norm": 0.8022708296775818, "learning_rate": 6.5725e-05, "loss": 0.1067, "step": 2630 }, { "grad_norm": 0.8659017086029053, "learning_rate": 6.5975e-05, "loss": 0.1109, "step": 2640 }, { "grad_norm": 0.9774784445762634, "learning_rate": 6.6225e-05, "loss": 0.1166, "step": 2650 }, { "grad_norm": 1.0638854503631592, "learning_rate": 6.6475e-05, "loss": 0.1111, "step": 2660 }, { "grad_norm": 0.9681568741798401, "learning_rate": 6.672500000000001e-05, "loss": 0.1253, "step": 2670 }, { "grad_norm": 1.1686567068099976, "learning_rate": 6.6975e-05, "loss": 0.124, "step": 2680 }, { "grad_norm": 0.814742386341095, "learning_rate": 6.722500000000001e-05, "loss": 0.1058, "step": 2690 }, { "grad_norm": 0.9554758667945862, "learning_rate": 6.7475e-05, "loss": 0.0932, "step": 2700 }, { "grad_norm": 0.6413518190383911, "learning_rate": 6.7725e-05, "loss": 0.1008, "step": 2710 }, { "grad_norm": 0.691417932510376, "learning_rate": 6.7975e-05, "loss": 0.1199, "step": 2720 }, { "grad_norm": 1.059353232383728, "learning_rate": 6.8225e-05, "loss": 0.1019, "step": 2730 }, { "grad_norm": 0.7214593291282654, "learning_rate": 6.8475e-05, "loss": 0.1035, "step": 2740 }, { "grad_norm": 0.7603000998497009, "learning_rate": 6.8725e-05, "loss": 0.1089, "step": 2750 }, { "grad_norm": 0.8605204224586487, "learning_rate": 6.8975e-05, "loss": 0.0935, "step": 2760 }, { "grad_norm": 0.7506943941116333, "learning_rate": 6.9225e-05, "loss": 0.0937, "step": 2770 }, { "grad_norm": 0.80689537525177, "learning_rate": 6.9475e-05, "loss": 0.1067, "step": 2780 }, { "grad_norm": 0.9976337552070618, "learning_rate": 6.9725e-05, "loss": 0.1069, "step": 2790 }, { "grad_norm": 0.7763229012489319, "learning_rate": 6.997500000000001e-05, "loss": 0.1041, "step": 2800 }, { "grad_norm": 1.43812096118927, "learning_rate": 7.022500000000001e-05, "loss": 0.1432, "step": 2810 }, { "grad_norm": 1.1177468299865723, "learning_rate": 7.0475e-05, "loss": 0.1163, "step": 2820 }, { "grad_norm": 0.6816694736480713, "learning_rate": 7.072500000000001e-05, "loss": 0.0925, "step": 2830 }, { "grad_norm": 0.7361778020858765, "learning_rate": 7.0975e-05, "loss": 0.123, "step": 2840 }, { "grad_norm": 0.7566016316413879, "learning_rate": 7.122500000000001e-05, "loss": 0.1223, "step": 2850 }, { "grad_norm": 0.8430366516113281, "learning_rate": 7.1475e-05, "loss": 0.1027, "step": 2860 }, { "grad_norm": 0.7344993948936462, "learning_rate": 7.172500000000001e-05, "loss": 0.11, "step": 2870 }, { "grad_norm": 0.793965756893158, "learning_rate": 7.1975e-05, "loss": 0.0934, "step": 2880 }, { "grad_norm": 0.8692274689674377, "learning_rate": 7.2225e-05, "loss": 0.1143, "step": 2890 }, { "grad_norm": 0.6707243919372559, "learning_rate": 7.2475e-05, "loss": 0.1004, "step": 2900 }, { "grad_norm": 1.0200694799423218, "learning_rate": 7.272499999999999e-05, "loss": 0.0959, "step": 2910 }, { "grad_norm": 0.5820983052253723, "learning_rate": 7.2975e-05, "loss": 0.1064, "step": 2920 }, { "grad_norm": 0.9550685882568359, "learning_rate": 7.3225e-05, "loss": 0.0897, "step": 2930 }, { "grad_norm": 0.6422669291496277, "learning_rate": 7.347500000000001e-05, "loss": 0.0831, "step": 2940 }, { "grad_norm": 0.6983286738395691, "learning_rate": 7.3725e-05, "loss": 0.0874, "step": 2950 }, { "grad_norm": 0.8769227862358093, "learning_rate": 7.397500000000001e-05, "loss": 0.1074, "step": 2960 }, { "grad_norm": 0.7716142535209656, "learning_rate": 7.4225e-05, "loss": 0.0886, "step": 2970 }, { "grad_norm": 0.9185315370559692, "learning_rate": 7.447500000000001e-05, "loss": 0.0899, "step": 2980 }, { "grad_norm": 0.7853766083717346, "learning_rate": 7.4725e-05, "loss": 0.1068, "step": 2990 }, { "grad_norm": 0.7476876974105835, "learning_rate": 7.4975e-05, "loss": 0.1047, "step": 3000 }, { "grad_norm": 0.8271065354347229, "learning_rate": 7.5225e-05, "loss": 0.0901, "step": 3010 }, { "grad_norm": 0.7498009204864502, "learning_rate": 7.5475e-05, "loss": 0.0931, "step": 3020 }, { "grad_norm": 0.6816725134849548, "learning_rate": 7.5725e-05, "loss": 0.0969, "step": 3030 }, { "grad_norm": 0.9351115822792053, "learning_rate": 7.5975e-05, "loss": 0.096, "step": 3040 }, { "grad_norm": 0.9783275127410889, "learning_rate": 7.6225e-05, "loss": 0.1099, "step": 3050 }, { "grad_norm": 0.8021042346954346, "learning_rate": 7.6475e-05, "loss": 0.0973, "step": 3060 }, { "grad_norm": 0.8193434476852417, "learning_rate": 7.672500000000001e-05, "loss": 0.0897, "step": 3070 }, { "grad_norm": 0.7028051614761353, "learning_rate": 7.697500000000001e-05, "loss": 0.0992, "step": 3080 }, { "grad_norm": 0.9103342294692993, "learning_rate": 7.722500000000001e-05, "loss": 0.1154, "step": 3090 }, { "grad_norm": 0.9772003293037415, "learning_rate": 7.747500000000001e-05, "loss": 0.1053, "step": 3100 }, { "grad_norm": 0.7669035792350769, "learning_rate": 7.7725e-05, "loss": 0.0956, "step": 3110 }, { "grad_norm": 0.9457706809043884, "learning_rate": 7.797500000000001e-05, "loss": 0.0881, "step": 3120 }, { "grad_norm": 1.036010980606079, "learning_rate": 7.8225e-05, "loss": 0.0889, "step": 3130 }, { "grad_norm": 0.7511409521102905, "learning_rate": 7.8475e-05, "loss": 0.1071, "step": 3140 }, { "grad_norm": 0.7153608202934265, "learning_rate": 7.8725e-05, "loss": 0.0931, "step": 3150 }, { "grad_norm": 0.795220673084259, "learning_rate": 7.8975e-05, "loss": 0.1034, "step": 3160 }, { "grad_norm": 0.6964222192764282, "learning_rate": 7.9225e-05, "loss": 0.1033, "step": 3170 }, { "grad_norm": 0.7678008675575256, "learning_rate": 7.9475e-05, "loss": 0.1053, "step": 3180 }, { "grad_norm": 0.8008227944374084, "learning_rate": 7.9725e-05, "loss": 0.1136, "step": 3190 }, { "grad_norm": 0.6288532018661499, "learning_rate": 7.9975e-05, "loss": 0.0959, "step": 3200 }, { "grad_norm": 0.6175469756126404, "learning_rate": 8.022500000000001e-05, "loss": 0.0997, "step": 3210 }, { "grad_norm": 0.9970079064369202, "learning_rate": 8.0475e-05, "loss": 0.1282, "step": 3220 }, { "grad_norm": 0.8363833427429199, "learning_rate": 8.072500000000001e-05, "loss": 0.1015, "step": 3230 }, { "grad_norm": 0.5861504673957825, "learning_rate": 8.0975e-05, "loss": 0.091, "step": 3240 }, { "grad_norm": 0.7049949169158936, "learning_rate": 8.122500000000001e-05, "loss": 0.0947, "step": 3250 }, { "grad_norm": 0.8293695449829102, "learning_rate": 8.1475e-05, "loss": 0.1, "step": 3260 }, { "grad_norm": 0.8599765300750732, "learning_rate": 8.172500000000001e-05, "loss": 0.0962, "step": 3270 }, { "grad_norm": 0.8503682613372803, "learning_rate": 8.1975e-05, "loss": 0.0965, "step": 3280 }, { "grad_norm": 0.7605143785476685, "learning_rate": 8.2225e-05, "loss": 0.084, "step": 3290 }, { "grad_norm": 0.7002519965171814, "learning_rate": 8.2475e-05, "loss": 0.0838, "step": 3300 }, { "grad_norm": 0.8034628033638, "learning_rate": 8.2725e-05, "loss": 0.1043, "step": 3310 }, { "grad_norm": 0.6916744709014893, "learning_rate": 8.2975e-05, "loss": 0.12, "step": 3320 }, { "grad_norm": 0.6446704864501953, "learning_rate": 8.3225e-05, "loss": 0.0998, "step": 3330 }, { "grad_norm": 1.0052764415740967, "learning_rate": 8.347500000000001e-05, "loss": 0.091, "step": 3340 }, { "grad_norm": 0.8108813762664795, "learning_rate": 8.3725e-05, "loss": 0.0984, "step": 3350 }, { "grad_norm": 1.0966055393218994, "learning_rate": 8.397500000000001e-05, "loss": 0.095, "step": 3360 }, { "grad_norm": 0.5920180082321167, "learning_rate": 8.422500000000001e-05, "loss": 0.1053, "step": 3370 }, { "grad_norm": 0.7883594036102295, "learning_rate": 8.447500000000001e-05, "loss": 0.1377, "step": 3380 }, { "grad_norm": 0.6832389831542969, "learning_rate": 8.4725e-05, "loss": 0.091, "step": 3390 }, { "grad_norm": 1.006657361984253, "learning_rate": 8.4975e-05, "loss": 0.1024, "step": 3400 }, { "grad_norm": 0.7356742024421692, "learning_rate": 8.5225e-05, "loss": 0.0994, "step": 3410 }, { "grad_norm": 0.8837589621543884, "learning_rate": 8.5475e-05, "loss": 0.0893, "step": 3420 }, { "grad_norm": 0.9948123097419739, "learning_rate": 8.5725e-05, "loss": 0.0933, "step": 3430 }, { "grad_norm": 0.6611402034759521, "learning_rate": 8.5975e-05, "loss": 0.1123, "step": 3440 }, { "grad_norm": 0.7010959982872009, "learning_rate": 8.6225e-05, "loss": 0.0945, "step": 3450 }, { "grad_norm": 0.6682335734367371, "learning_rate": 8.6475e-05, "loss": 0.0876, "step": 3460 }, { "grad_norm": 0.517678439617157, "learning_rate": 8.672500000000001e-05, "loss": 0.1103, "step": 3470 }, { "grad_norm": 0.7084426879882812, "learning_rate": 8.6975e-05, "loss": 0.0997, "step": 3480 }, { "grad_norm": 0.8854421377182007, "learning_rate": 8.7225e-05, "loss": 0.0946, "step": 3490 }, { "grad_norm": 0.8852643966674805, "learning_rate": 8.747500000000001e-05, "loss": 0.1023, "step": 3500 }, { "grad_norm": 0.8449098467826843, "learning_rate": 8.7725e-05, "loss": 0.1059, "step": 3510 }, { "grad_norm": 0.6406853795051575, "learning_rate": 8.797500000000001e-05, "loss": 0.0895, "step": 3520 }, { "grad_norm": 0.9077081084251404, "learning_rate": 8.8225e-05, "loss": 0.0982, "step": 3530 }, { "grad_norm": 0.6679278016090393, "learning_rate": 8.847500000000001e-05, "loss": 0.0921, "step": 3540 }, { "grad_norm": 0.6262614727020264, "learning_rate": 8.8725e-05, "loss": 0.0873, "step": 3550 }, { "grad_norm": 0.5614915490150452, "learning_rate": 8.897500000000001e-05, "loss": 0.0766, "step": 3560 }, { "grad_norm": 0.6095160245895386, "learning_rate": 8.9225e-05, "loss": 0.1005, "step": 3570 }, { "grad_norm": 0.9844671487808228, "learning_rate": 8.9475e-05, "loss": 0.11, "step": 3580 }, { "grad_norm": 0.71759033203125, "learning_rate": 8.9725e-05, "loss": 0.1046, "step": 3590 }, { "grad_norm": 0.7496538758277893, "learning_rate": 8.9975e-05, "loss": 0.1033, "step": 3600 }, { "grad_norm": 0.704156219959259, "learning_rate": 9.0225e-05, "loss": 0.0949, "step": 3610 }, { "grad_norm": 0.6249984502792358, "learning_rate": 9.0475e-05, "loss": 0.0955, "step": 3620 }, { "grad_norm": 0.6239193677902222, "learning_rate": 9.072500000000001e-05, "loss": 0.0832, "step": 3630 }, { "grad_norm": 0.6589145064353943, "learning_rate": 9.0975e-05, "loss": 0.0922, "step": 3640 }, { "grad_norm": 0.8034786581993103, "learning_rate": 9.122500000000001e-05, "loss": 0.1041, "step": 3650 }, { "grad_norm": 0.7687738537788391, "learning_rate": 9.1475e-05, "loss": 0.0915, "step": 3660 }, { "grad_norm": 0.7347885966300964, "learning_rate": 9.172500000000001e-05, "loss": 0.0987, "step": 3670 }, { "grad_norm": 0.719623863697052, "learning_rate": 9.1975e-05, "loss": 0.0979, "step": 3680 }, { "grad_norm": 0.8473547697067261, "learning_rate": 9.2225e-05, "loss": 0.0814, "step": 3690 }, { "grad_norm": 0.8008864521980286, "learning_rate": 9.2475e-05, "loss": 0.0978, "step": 3700 }, { "grad_norm": 0.7939205169677734, "learning_rate": 9.2725e-05, "loss": 0.094, "step": 3710 }, { "grad_norm": 0.8476489782333374, "learning_rate": 9.2975e-05, "loss": 0.0979, "step": 3720 }, { "grad_norm": 0.5727393627166748, "learning_rate": 9.3225e-05, "loss": 0.0821, "step": 3730 }, { "grad_norm": 0.5411539077758789, "learning_rate": 9.3475e-05, "loss": 0.0777, "step": 3740 }, { "grad_norm": 0.7021387219429016, "learning_rate": 9.3725e-05, "loss": 0.0867, "step": 3750 }, { "grad_norm": 0.542326807975769, "learning_rate": 9.397500000000001e-05, "loss": 0.0722, "step": 3760 }, { "grad_norm": 0.7896072268486023, "learning_rate": 9.422500000000001e-05, "loss": 0.0987, "step": 3770 }, { "grad_norm": 0.7487494945526123, "learning_rate": 9.4475e-05, "loss": 0.1096, "step": 3780 }, { "grad_norm": 0.6207165122032166, "learning_rate": 9.472500000000001e-05, "loss": 0.0849, "step": 3790 }, { "grad_norm": 0.6570515632629395, "learning_rate": 9.4975e-05, "loss": 0.1142, "step": 3800 }, { "grad_norm": 0.8968174457550049, "learning_rate": 9.522500000000001e-05, "loss": 0.0821, "step": 3810 }, { "grad_norm": 0.9383790493011475, "learning_rate": 9.5475e-05, "loss": 0.1165, "step": 3820 }, { "grad_norm": 0.725584089756012, "learning_rate": 9.572500000000001e-05, "loss": 0.0846, "step": 3830 }, { "grad_norm": 0.601539671421051, "learning_rate": 9.5975e-05, "loss": 0.0762, "step": 3840 }, { "grad_norm": 0.647935152053833, "learning_rate": 9.622500000000001e-05, "loss": 0.0958, "step": 3850 }, { "grad_norm": 0.7970088124275208, "learning_rate": 9.6475e-05, "loss": 0.078, "step": 3860 }, { "grad_norm": 0.7625325918197632, "learning_rate": 9.6725e-05, "loss": 0.0892, "step": 3870 }, { "grad_norm": 0.5277842879295349, "learning_rate": 9.6975e-05, "loss": 0.0741, "step": 3880 }, { "grad_norm": 0.9385785460472107, "learning_rate": 9.7225e-05, "loss": 0.097, "step": 3890 }, { "grad_norm": 0.5395745038986206, "learning_rate": 9.747500000000001e-05, "loss": 0.0903, "step": 3900 }, { "grad_norm": 0.6532346606254578, "learning_rate": 9.7725e-05, "loss": 0.0942, "step": 3910 }, { "grad_norm": 0.8373002409934998, "learning_rate": 9.797500000000001e-05, "loss": 0.1049, "step": 3920 }, { "grad_norm": 0.5717490911483765, "learning_rate": 9.8225e-05, "loss": 0.0808, "step": 3930 }, { "grad_norm": 0.6229724884033203, "learning_rate": 9.847500000000001e-05, "loss": 0.1144, "step": 3940 }, { "grad_norm": 0.731971800327301, "learning_rate": 9.8725e-05, "loss": 0.1025, "step": 3950 }, { "grad_norm": 0.6094070076942444, "learning_rate": 9.897500000000001e-05, "loss": 0.1068, "step": 3960 }, { "grad_norm": 0.9418469071388245, "learning_rate": 9.9225e-05, "loss": 0.0882, "step": 3970 }, { "grad_norm": 0.5513692498207092, "learning_rate": 9.9475e-05, "loss": 0.0993, "step": 3980 }, { "grad_norm": 0.7648428082466125, "learning_rate": 9.9725e-05, "loss": 0.1095, "step": 3990 }, { "grad_norm": 0.9381961822509766, "learning_rate": 9.9975e-05, "loss": 0.0905, "step": 4000 }, { "grad_norm": 0.6357324123382568, "learning_rate": 9.999999653982884e-05, "loss": 0.0876, "step": 4010 }, { "grad_norm": 0.5407594442367554, "learning_rate": 9.999998457874392e-05, "loss": 0.0951, "step": 4020 }, { "grad_norm": 0.701244592666626, "learning_rate": 9.999996407402913e-05, "loss": 0.0954, "step": 4030 }, { "grad_norm": 0.6285465359687805, "learning_rate": 9.999993502568801e-05, "loss": 0.0725, "step": 4040 }, { "grad_norm": 0.6869670748710632, "learning_rate": 9.999989743372548e-05, "loss": 0.093, "step": 4050 }, { "grad_norm": 0.5623934268951416, "learning_rate": 9.999985129814798e-05, "loss": 0.0916, "step": 4060 }, { "grad_norm": 0.6948279142379761, "learning_rate": 9.99997966189634e-05, "loss": 0.0901, "step": 4070 }, { "grad_norm": 0.5484039187431335, "learning_rate": 9.999973339618107e-05, "loss": 0.1024, "step": 4080 }, { "grad_norm": 0.5890176892280579, "learning_rate": 9.999966162981179e-05, "loss": 0.1017, "step": 4090 }, { "grad_norm": 0.6861647367477417, "learning_rate": 9.999958131986784e-05, "loss": 0.0914, "step": 4100 }, { "grad_norm": 0.8814710974693298, "learning_rate": 9.999949246636293e-05, "loss": 0.0936, "step": 4110 }, { "grad_norm": 0.8523208498954773, "learning_rate": 9.999939506931224e-05, "loss": 0.1038, "step": 4120 }, { "grad_norm": 0.7485443353652954, "learning_rate": 9.999928912873243e-05, "loss": 0.0918, "step": 4130 }, { "grad_norm": 0.6947469711303711, "learning_rate": 9.999917464464159e-05, "loss": 0.0975, "step": 4140 }, { "grad_norm": 0.6469029188156128, "learning_rate": 9.999905161705929e-05, "loss": 0.1072, "step": 4150 }, { "grad_norm": 0.6330101490020752, "learning_rate": 9.999892004600653e-05, "loss": 0.081, "step": 4160 }, { "grad_norm": 0.55695641040802, "learning_rate": 9.999877993150581e-05, "loss": 0.0738, "step": 4170 }, { "grad_norm": 0.7032655477523804, "learning_rate": 9.999863127358108e-05, "loss": 0.0745, "step": 4180 }, { "grad_norm": 0.5208277106285095, "learning_rate": 9.999847407225773e-05, "loss": 0.0758, "step": 4190 }, { "grad_norm": 0.5705861449241638, "learning_rate": 9.999830832756262e-05, "loss": 0.0799, "step": 4200 }, { "grad_norm": 0.5744884610176086, "learning_rate": 9.999813403952407e-05, "loss": 0.0833, "step": 4210 }, { "grad_norm": 0.7418793439865112, "learning_rate": 9.999795120817187e-05, "loss": 0.0881, "step": 4220 }, { "grad_norm": 0.6645041108131409, "learning_rate": 9.999775983353725e-05, "loss": 0.1066, "step": 4230 }, { "grad_norm": 0.5176652669906616, "learning_rate": 9.999755991565292e-05, "loss": 0.1012, "step": 4240 }, { "grad_norm": 1.0214883089065552, "learning_rate": 9.999735145455303e-05, "loss": 0.1081, "step": 4250 }, { "grad_norm": 0.5727677941322327, "learning_rate": 9.99971344502732e-05, "loss": 0.1149, "step": 4260 }, { "grad_norm": 0.6400360465049744, "learning_rate": 9.999690890285053e-05, "loss": 0.0952, "step": 4270 }, { "grad_norm": 0.7398287057876587, "learning_rate": 9.999667481232356e-05, "loss": 0.0885, "step": 4280 }, { "grad_norm": 0.6649957299232483, "learning_rate": 9.999643217873225e-05, "loss": 0.0897, "step": 4290 }, { "grad_norm": 0.7372796535491943, "learning_rate": 9.999618100211809e-05, "loss": 0.0744, "step": 4300 }, { "grad_norm": 0.5514634251594543, "learning_rate": 9.999592128252402e-05, "loss": 0.0994, "step": 4310 }, { "grad_norm": 0.630398690700531, "learning_rate": 9.999565301999437e-05, "loss": 0.1096, "step": 4320 }, { "grad_norm": 0.5649704337120056, "learning_rate": 9.999537621457502e-05, "loss": 0.0748, "step": 4330 }, { "grad_norm": 0.6740155220031738, "learning_rate": 9.999509086631323e-05, "loss": 0.0734, "step": 4340 }, { "grad_norm": 0.519449770450592, "learning_rate": 9.99947969752578e-05, "loss": 0.0734, "step": 4350 }, { "grad_norm": 0.6989755034446716, "learning_rate": 9.999449454145891e-05, "loss": 0.0823, "step": 4360 }, { "grad_norm": 0.8703180551528931, "learning_rate": 9.999418356496827e-05, "loss": 0.0801, "step": 4370 }, { "grad_norm": 0.5367271304130554, "learning_rate": 9.999386404583899e-05, "loss": 0.0949, "step": 4380 }, { "grad_norm": 0.6826335191726685, "learning_rate": 9.999353598412568e-05, "loss": 0.0853, "step": 4390 }, { "grad_norm": 0.5789029002189636, "learning_rate": 9.999319937988442e-05, "loss": 0.0846, "step": 4400 }, { "grad_norm": 0.527130126953125, "learning_rate": 9.999285423317268e-05, "loss": 0.0827, "step": 4410 }, { "grad_norm": 0.5768059492111206, "learning_rate": 9.999250054404947e-05, "loss": 0.1004, "step": 4420 }, { "grad_norm": 0.5687564611434937, "learning_rate": 9.99921383125752e-05, "loss": 0.0686, "step": 4430 }, { "grad_norm": 0.5402534008026123, "learning_rate": 9.99917675388118e-05, "loss": 0.0968, "step": 4440 }, { "grad_norm": 0.8411921858787537, "learning_rate": 9.99913882228226e-05, "loss": 0.1037, "step": 4450 }, { "grad_norm": 0.5587680339813232, "learning_rate": 9.999100036467242e-05, "loss": 0.0866, "step": 4460 }, { "grad_norm": 0.5041693449020386, "learning_rate": 9.999060396442753e-05, "loss": 0.0835, "step": 4470 }, { "grad_norm": 0.5602508187294006, "learning_rate": 9.999019902215566e-05, "loss": 0.0997, "step": 4480 }, { "grad_norm": 0.6382730007171631, "learning_rate": 9.998978553792602e-05, "loss": 0.1038, "step": 4490 }, { "grad_norm": 0.6599559187889099, "learning_rate": 9.998936351180926e-05, "loss": 0.0794, "step": 4500 }, { "grad_norm": 0.6940817832946777, "learning_rate": 9.998893294387747e-05, "loss": 0.1098, "step": 4510 }, { "grad_norm": 0.7730697393417358, "learning_rate": 9.998849383420426e-05, "loss": 0.0873, "step": 4520 }, { "grad_norm": 0.5478499531745911, "learning_rate": 9.998804618286465e-05, "loss": 0.0887, "step": 4530 }, { "grad_norm": 1.0400426387786865, "learning_rate": 9.99875899899351e-05, "loss": 0.1136, "step": 4540 }, { "grad_norm": 0.6939018964767456, "learning_rate": 9.99871252554936e-05, "loss": 0.0999, "step": 4550 }, { "grad_norm": 0.5208152532577515, "learning_rate": 9.998665197961955e-05, "loss": 0.0825, "step": 4560 }, { "grad_norm": 0.7848846912384033, "learning_rate": 9.998617016239379e-05, "loss": 0.111, "step": 4570 }, { "grad_norm": 0.9645605087280273, "learning_rate": 9.998567980389869e-05, "loss": 0.1006, "step": 4580 }, { "grad_norm": 0.6650351285934448, "learning_rate": 9.998518090421802e-05, "loss": 0.094, "step": 4590 }, { "grad_norm": 0.6574291586875916, "learning_rate": 9.998467346343703e-05, "loss": 0.0866, "step": 4600 }, { "grad_norm": 0.5688109993934631, "learning_rate": 9.998415748164243e-05, "loss": 0.0829, "step": 4610 }, { "grad_norm": 0.6697635650634766, "learning_rate": 9.998363295892238e-05, "loss": 0.0887, "step": 4620 }, { "grad_norm": 0.8371644616127014, "learning_rate": 9.998309989536652e-05, "loss": 0.1044, "step": 4630 }, { "grad_norm": 0.6755525469779968, "learning_rate": 9.998255829106593e-05, "loss": 0.098, "step": 4640 }, { "grad_norm": 0.8626416921615601, "learning_rate": 9.998200814611316e-05, "loss": 0.0869, "step": 4650 }, { "grad_norm": 0.6151751279830933, "learning_rate": 9.998144946060219e-05, "loss": 0.0902, "step": 4660 }, { "grad_norm": 0.8522558212280273, "learning_rate": 9.998088223462852e-05, "loss": 0.0816, "step": 4670 }, { "grad_norm": 0.5120728015899658, "learning_rate": 9.998030646828905e-05, "loss": 0.0758, "step": 4680 }, { "grad_norm": 0.5669979453086853, "learning_rate": 9.997972216168217e-05, "loss": 0.0884, "step": 4690 }, { "grad_norm": 0.703789234161377, "learning_rate": 9.997912931490771e-05, "loss": 0.0808, "step": 4700 }, { "grad_norm": 0.6049168109893799, "learning_rate": 9.9978527928067e-05, "loss": 0.0815, "step": 4710 }, { "grad_norm": 0.601828396320343, "learning_rate": 9.997791800126277e-05, "loss": 0.0782, "step": 4720 }, { "grad_norm": 0.7571966052055359, "learning_rate": 9.997729953459927e-05, "loss": 0.1088, "step": 4730 }, { "grad_norm": 0.5689601302146912, "learning_rate": 9.997667252818214e-05, "loss": 0.092, "step": 4740 }, { "grad_norm": 0.5209698677062988, "learning_rate": 9.997603698211855e-05, "loss": 0.075, "step": 4750 }, { "grad_norm": 0.694490909576416, "learning_rate": 9.99753928965171e-05, "loss": 0.0934, "step": 4760 }, { "grad_norm": 0.6790400147438049, "learning_rate": 9.997474027148781e-05, "loss": 0.0925, "step": 4770 }, { "grad_norm": 0.9209374189376831, "learning_rate": 9.997407910714223e-05, "loss": 0.1068, "step": 4780 }, { "grad_norm": 0.5803106427192688, "learning_rate": 9.997340940359332e-05, "loss": 0.0756, "step": 4790 }, { "grad_norm": 0.5994899868965149, "learning_rate": 9.997273116095552e-05, "loss": 0.0782, "step": 4800 }, { "grad_norm": 0.5473602414131165, "learning_rate": 9.997204437934473e-05, "loss": 0.0863, "step": 4810 }, { "grad_norm": 0.712099552154541, "learning_rate": 9.997134905887829e-05, "loss": 0.0923, "step": 4820 }, { "grad_norm": 0.5876806378364563, "learning_rate": 9.997064519967501e-05, "loss": 0.0805, "step": 4830 }, { "grad_norm": 0.5629193782806396, "learning_rate": 9.996993280185517e-05, "loss": 0.0962, "step": 4840 }, { "grad_norm": 0.6138964295387268, "learning_rate": 9.99692118655405e-05, "loss": 0.0811, "step": 4850 }, { "grad_norm": 0.6555774807929993, "learning_rate": 9.996848239085417e-05, "loss": 0.0826, "step": 4860 }, { "grad_norm": 0.7674108147621155, "learning_rate": 9.996774437792085e-05, "loss": 0.0858, "step": 4870 }, { "grad_norm": 0.5726824998855591, "learning_rate": 9.996699782686664e-05, "loss": 0.0765, "step": 4880 }, { "grad_norm": 0.9793755412101746, "learning_rate": 9.996624273781909e-05, "loss": 0.0982, "step": 4890 }, { "grad_norm": 1.0968393087387085, "learning_rate": 9.996547911090725e-05, "loss": 0.0933, "step": 4900 }, { "grad_norm": 0.6638041734695435, "learning_rate": 9.996470694626157e-05, "loss": 0.0984, "step": 4910 }, { "grad_norm": 0.5429790019989014, "learning_rate": 9.996392624401403e-05, "loss": 0.0999, "step": 4920 }, { "grad_norm": 0.5987622141838074, "learning_rate": 9.996313700429801e-05, "loss": 0.0933, "step": 4930 }, { "grad_norm": 0.6047362089157104, "learning_rate": 9.996233922724836e-05, "loss": 0.0845, "step": 4940 }, { "grad_norm": 0.6244322657585144, "learning_rate": 9.996153291300141e-05, "loss": 0.0909, "step": 4950 }, { "grad_norm": 0.44777339696884155, "learning_rate": 9.996071806169494e-05, "loss": 0.0739, "step": 4960 }, { "grad_norm": 0.6200783252716064, "learning_rate": 9.995989467346817e-05, "loss": 0.0892, "step": 4970 }, { "grad_norm": 0.5750377178192139, "learning_rate": 9.995906274846183e-05, "loss": 0.0912, "step": 4980 }, { "grad_norm": 0.7204052209854126, "learning_rate": 9.995822228681803e-05, "loss": 0.0847, "step": 4990 }, { "grad_norm": 0.6802074313163757, "learning_rate": 9.99573732886804e-05, "loss": 0.0973, "step": 5000 }, { "grad_norm": 0.5200930833816528, "learning_rate": 9.995651575419402e-05, "loss": 0.081, "step": 5010 }, { "grad_norm": 0.6946431994438171, "learning_rate": 9.995564968350541e-05, "loss": 0.086, "step": 5020 }, { "grad_norm": 0.49099817872047424, "learning_rate": 9.995477507676256e-05, "loss": 0.0716, "step": 5030 }, { "grad_norm": 0.4977186620235443, "learning_rate": 9.995389193411493e-05, "loss": 0.0743, "step": 5040 }, { "grad_norm": 0.6478691101074219, "learning_rate": 9.995300025571339e-05, "loss": 0.0789, "step": 5050 }, { "grad_norm": 0.5447460412979126, "learning_rate": 9.995210004171034e-05, "loss": 0.0591, "step": 5060 }, { "grad_norm": 0.6369962096214294, "learning_rate": 9.995119129225956e-05, "loss": 0.0964, "step": 5070 }, { "grad_norm": 0.9290935397148132, "learning_rate": 9.995027400751637e-05, "loss": 0.0873, "step": 5080 }, { "grad_norm": 0.6302912831306458, "learning_rate": 9.994934818763751e-05, "loss": 0.0754, "step": 5090 }, { "grad_norm": 0.5011770129203796, "learning_rate": 9.994841383278115e-05, "loss": 0.0759, "step": 5100 }, { "grad_norm": 0.5423506498336792, "learning_rate": 9.994747094310695e-05, "loss": 0.0797, "step": 5110 }, { "grad_norm": 0.5638033151626587, "learning_rate": 9.994651951877604e-05, "loss": 0.0914, "step": 5120 }, { "grad_norm": 0.5683214068412781, "learning_rate": 9.994555955995099e-05, "loss": 0.0869, "step": 5130 }, { "grad_norm": 0.6378177404403687, "learning_rate": 9.994459106679581e-05, "loss": 0.0769, "step": 5140 }, { "grad_norm": 0.4721810817718506, "learning_rate": 9.994361403947603e-05, "loss": 0.068, "step": 5150 }, { "grad_norm": 0.620423436164856, "learning_rate": 9.994262847815854e-05, "loss": 0.0904, "step": 5160 }, { "grad_norm": 0.6360698342323303, "learning_rate": 9.99416343830118e-05, "loss": 0.0718, "step": 5170 }, { "grad_norm": 0.4958604574203491, "learning_rate": 9.994063175420565e-05, "loss": 0.0653, "step": 5180 }, { "grad_norm": 0.6619305610656738, "learning_rate": 9.99396205919114e-05, "loss": 0.0889, "step": 5190 }, { "grad_norm": 0.6328627467155457, "learning_rate": 9.993860089630185e-05, "loss": 0.0642, "step": 5200 }, { "grad_norm": 0.7248746752738953, "learning_rate": 9.993757266755123e-05, "loss": 0.0759, "step": 5210 }, { "grad_norm": 0.6326178908348083, "learning_rate": 9.993653590583522e-05, "loss": 0.0792, "step": 5220 }, { "grad_norm": 0.6612693071365356, "learning_rate": 9.993549061133102e-05, "loss": 0.0762, "step": 5230 }, { "grad_norm": 0.4331805408000946, "learning_rate": 9.993443678421719e-05, "loss": 0.0689, "step": 5240 }, { "grad_norm": 0.5074344873428345, "learning_rate": 9.993337442467384e-05, "loss": 0.0892, "step": 5250 }, { "grad_norm": 0.6849880218505859, "learning_rate": 9.993230353288248e-05, "loss": 0.0678, "step": 5260 }, { "grad_norm": 0.6853891015052795, "learning_rate": 9.993122410902608e-05, "loss": 0.0803, "step": 5270 }, { "grad_norm": 0.6720630526542664, "learning_rate": 9.993013615328912e-05, "loss": 0.0936, "step": 5280 }, { "grad_norm": 0.5126381516456604, "learning_rate": 9.992903966585747e-05, "loss": 0.0901, "step": 5290 }, { "grad_norm": 0.5445653796195984, "learning_rate": 9.992793464691852e-05, "loss": 0.0812, "step": 5300 }, { "grad_norm": 0.673541247844696, "learning_rate": 9.992682109666105e-05, "loss": 0.0651, "step": 5310 }, { "grad_norm": 0.5698822736740112, "learning_rate": 9.992569901527538e-05, "loss": 0.0801, "step": 5320 }, { "grad_norm": 0.6669121980667114, "learning_rate": 9.99245684029532e-05, "loss": 0.0936, "step": 5330 }, { "grad_norm": 0.5372901558876038, "learning_rate": 9.992342925988774e-05, "loss": 0.0951, "step": 5340 }, { "grad_norm": 0.44083118438720703, "learning_rate": 9.992228158627361e-05, "loss": 0.0701, "step": 5350 }, { "grad_norm": 0.6734235882759094, "learning_rate": 9.992112538230693e-05, "loss": 0.0798, "step": 5360 }, { "grad_norm": 0.4557068943977356, "learning_rate": 9.991996064818527e-05, "loss": 0.0576, "step": 5370 }, { "grad_norm": 0.6037994623184204, "learning_rate": 9.991878738410768e-05, "loss": 0.0706, "step": 5380 }, { "grad_norm": 0.4992116093635559, "learning_rate": 9.991760559027457e-05, "loss": 0.0727, "step": 5390 }, { "grad_norm": 0.5690847635269165, "learning_rate": 9.991641526688793e-05, "loss": 0.0836, "step": 5400 }, { "grad_norm": 0.5494564175605774, "learning_rate": 9.991521641415113e-05, "loss": 0.0625, "step": 5410 }, { "grad_norm": 0.5549644827842712, "learning_rate": 9.991400903226904e-05, "loss": 0.0706, "step": 5420 }, { "grad_norm": 0.8471283316612244, "learning_rate": 9.991279312144794e-05, "loss": 0.0757, "step": 5430 }, { "grad_norm": 0.5045434832572937, "learning_rate": 9.991156868189564e-05, "loss": 0.0862, "step": 5440 }, { "grad_norm": 0.6581383347511292, "learning_rate": 9.991033571382131e-05, "loss": 0.0609, "step": 5450 }, { "grad_norm": 0.6963561177253723, "learning_rate": 9.990909421743569e-05, "loss": 0.0686, "step": 5460 }, { "grad_norm": 0.5752547979354858, "learning_rate": 9.990784419295085e-05, "loss": 0.0681, "step": 5470 }, { "grad_norm": 0.554439902305603, "learning_rate": 9.990658564058044e-05, "loss": 0.0833, "step": 5480 }, { "grad_norm": 0.49191370606422424, "learning_rate": 9.990531856053948e-05, "loss": 0.0761, "step": 5490 }, { "grad_norm": 0.6112287640571594, "learning_rate": 9.99040429530445e-05, "loss": 0.0822, "step": 5500 }, { "grad_norm": 0.6559996008872986, "learning_rate": 9.990275881831346e-05, "loss": 0.074, "step": 5510 }, { "grad_norm": 0.5800774693489075, "learning_rate": 9.990146615656577e-05, "loss": 0.061, "step": 5520 }, { "grad_norm": 0.5424721240997314, "learning_rate": 9.990016496802233e-05, "loss": 0.0673, "step": 5530 }, { "grad_norm": 0.5157883167266846, "learning_rate": 9.989885525290548e-05, "loss": 0.0683, "step": 5540 }, { "grad_norm": 0.5363186597824097, "learning_rate": 9.989753701143897e-05, "loss": 0.0722, "step": 5550 }, { "grad_norm": 0.5072373151779175, "learning_rate": 9.989621024384812e-05, "loss": 0.0738, "step": 5560 }, { "grad_norm": 0.559445858001709, "learning_rate": 9.989487495035959e-05, "loss": 0.0709, "step": 5570 }, { "grad_norm": 0.7315024137496948, "learning_rate": 9.989353113120156e-05, "loss": 0.0997, "step": 5580 }, { "grad_norm": 0.5915868878364563, "learning_rate": 9.989217878660366e-05, "loss": 0.0658, "step": 5590 }, { "grad_norm": 0.7162283658981323, "learning_rate": 9.989081791679695e-05, "loss": 0.0909, "step": 5600 }, { "grad_norm": 0.5756527781486511, "learning_rate": 9.988944852201397e-05, "loss": 0.0755, "step": 5610 }, { "grad_norm": 0.5860406160354614, "learning_rate": 9.988807060248873e-05, "loss": 0.0748, "step": 5620 }, { "grad_norm": 0.5014796257019043, "learning_rate": 9.988668415845665e-05, "loss": 0.0651, "step": 5630 }, { "grad_norm": 0.6094543933868408, "learning_rate": 9.988528919015466e-05, "loss": 0.0686, "step": 5640 }, { "grad_norm": 0.4859192669391632, "learning_rate": 9.988388569782112e-05, "loss": 0.0818, "step": 5650 }, { "grad_norm": 0.5417901277542114, "learning_rate": 9.988247368169583e-05, "loss": 0.073, "step": 5660 }, { "grad_norm": 0.7339745759963989, "learning_rate": 9.988105314202007e-05, "loss": 0.0825, "step": 5670 }, { "grad_norm": 0.6656520962715149, "learning_rate": 9.987962407903659e-05, "loss": 0.0886, "step": 5680 }, { "grad_norm": 0.6197206377983093, "learning_rate": 9.987818649298957e-05, "loss": 0.0837, "step": 5690 }, { "grad_norm": 0.6105799078941345, "learning_rate": 9.987674038412465e-05, "loss": 0.0777, "step": 5700 }, { "grad_norm": 0.4706457555294037, "learning_rate": 9.987528575268891e-05, "loss": 0.089, "step": 5710 }, { "grad_norm": 0.7284520268440247, "learning_rate": 9.987382259893095e-05, "loss": 0.0809, "step": 5720 }, { "grad_norm": 0.5611932277679443, "learning_rate": 9.987235092310074e-05, "loss": 0.0781, "step": 5730 }, { "grad_norm": 0.5631946325302124, "learning_rate": 9.987087072544978e-05, "loss": 0.0934, "step": 5740 }, { "grad_norm": 0.45742470026016235, "learning_rate": 9.9869382006231e-05, "loss": 0.0778, "step": 5750 }, { "grad_norm": 0.5659241676330566, "learning_rate": 9.986788476569875e-05, "loss": 0.08, "step": 5760 }, { "grad_norm": 0.5409712195396423, "learning_rate": 9.986637900410887e-05, "loss": 0.0836, "step": 5770 }, { "grad_norm": 0.6325596570968628, "learning_rate": 9.986486472171869e-05, "loss": 0.087, "step": 5780 }, { "grad_norm": 0.6824360489845276, "learning_rate": 9.986334191878692e-05, "loss": 0.0747, "step": 5790 }, { "grad_norm": 0.752435564994812, "learning_rate": 9.986181059557378e-05, "loss": 0.0869, "step": 5800 }, { "grad_norm": 0.6147380471229553, "learning_rate": 9.986027075234094e-05, "loss": 0.075, "step": 5810 }, { "grad_norm": 0.6780104041099548, "learning_rate": 9.985872238935152e-05, "loss": 0.0842, "step": 5820 }, { "grad_norm": 0.6511956453323364, "learning_rate": 9.985716550687008e-05, "loss": 0.0891, "step": 5830 }, { "grad_norm": 0.4721926152706146, "learning_rate": 9.985560010516264e-05, "loss": 0.0751, "step": 5840 }, { "grad_norm": 0.6878783702850342, "learning_rate": 9.985402618449668e-05, "loss": 0.073, "step": 5850 }, { "grad_norm": 0.6476760506629944, "learning_rate": 9.985244374514118e-05, "loss": 0.097, "step": 5860 }, { "grad_norm": 0.9828624129295349, "learning_rate": 9.985085278736651e-05, "loss": 0.095, "step": 5870 }, { "grad_norm": 0.5744736790657043, "learning_rate": 9.984925331144452e-05, "loss": 0.1004, "step": 5880 }, { "grad_norm": 0.42572781443595886, "learning_rate": 9.984764531764851e-05, "loss": 0.0738, "step": 5890 }, { "grad_norm": 0.5463350415229797, "learning_rate": 9.984602880625326e-05, "loss": 0.0979, "step": 5900 }, { "grad_norm": 0.4999721944332123, "learning_rate": 9.9844403777535e-05, "loss": 0.0882, "step": 5910 }, { "grad_norm": 0.8627490997314453, "learning_rate": 9.984277023177135e-05, "loss": 0.0783, "step": 5920 }, { "grad_norm": 0.5939971804618835, "learning_rate": 9.984112816924148e-05, "loss": 0.0626, "step": 5930 }, { "grad_norm": 0.7757953405380249, "learning_rate": 9.983947759022596e-05, "loss": 0.0909, "step": 5940 }, { "grad_norm": 0.5407822132110596, "learning_rate": 9.983781849500682e-05, "loss": 0.0715, "step": 5950 }, { "grad_norm": 0.5867881178855896, "learning_rate": 9.98361508838676e-05, "loss": 0.0688, "step": 5960 }, { "grad_norm": 0.6831774115562439, "learning_rate": 9.98344747570932e-05, "loss": 0.0771, "step": 5970 }, { "grad_norm": 0.5860695242881775, "learning_rate": 9.983279011497004e-05, "loss": 0.0721, "step": 5980 }, { "grad_norm": 0.775460958480835, "learning_rate": 9.983109695778596e-05, "loss": 0.0941, "step": 5990 }, { "grad_norm": 0.6097055077552795, "learning_rate": 9.982939528583032e-05, "loss": 0.0903, "step": 6000 }, { "grad_norm": 0.933905303478241, "learning_rate": 9.982768509939385e-05, "loss": 0.0802, "step": 6010 }, { "grad_norm": 0.5476328134536743, "learning_rate": 9.982596639876879e-05, "loss": 0.068, "step": 6020 }, { "grad_norm": 0.48217517137527466, "learning_rate": 9.982423918424881e-05, "loss": 0.0835, "step": 6030 }, { "grad_norm": 0.5323188900947571, "learning_rate": 9.982250345612908e-05, "loss": 0.0721, "step": 6040 }, { "grad_norm": 0.5656123757362366, "learning_rate": 9.982075921470611e-05, "loss": 0.0725, "step": 6050 }, { "grad_norm": 0.6056115031242371, "learning_rate": 9.981900646027802e-05, "loss": 0.0767, "step": 6060 }, { "grad_norm": 0.5441111922264099, "learning_rate": 9.981724519314425e-05, "loss": 0.0709, "step": 6070 }, { "grad_norm": 0.669802188873291, "learning_rate": 9.981547541360581e-05, "loss": 0.0836, "step": 6080 }, { "grad_norm": 0.5113499760627747, "learning_rate": 9.981369712196508e-05, "loss": 0.0685, "step": 6090 }, { "grad_norm": 0.5157586336135864, "learning_rate": 9.981191031852592e-05, "loss": 0.099, "step": 6100 }, { "grad_norm": 0.5737616419792175, "learning_rate": 9.981011500359362e-05, "loss": 0.0708, "step": 6110 }, { "grad_norm": 0.5606657266616821, "learning_rate": 9.9808311177475e-05, "loss": 0.0721, "step": 6120 }, { "grad_norm": 0.5593275427818298, "learning_rate": 9.980649884047826e-05, "loss": 0.0649, "step": 6130 }, { "grad_norm": 0.4755437672138214, "learning_rate": 9.980467799291307e-05, "loss": 0.0765, "step": 6140 }, { "grad_norm": 0.49631959199905396, "learning_rate": 9.980284863509058e-05, "loss": 0.0985, "step": 6150 }, { "grad_norm": 0.7180445790290833, "learning_rate": 9.980101076732334e-05, "loss": 0.1015, "step": 6160 }, { "grad_norm": 0.6729334592819214, "learning_rate": 9.979916438992544e-05, "loss": 0.0796, "step": 6170 }, { "grad_norm": 0.5663818120956421, "learning_rate": 9.979730950321237e-05, "loss": 0.0755, "step": 6180 }, { "grad_norm": 0.42042988538742065, "learning_rate": 9.979544610750104e-05, "loss": 0.0869, "step": 6190 }, { "grad_norm": 0.5069091320037842, "learning_rate": 9.97935742031099e-05, "loss": 0.0707, "step": 6200 }, { "grad_norm": 0.6903197169303894, "learning_rate": 9.979169379035878e-05, "loss": 0.0817, "step": 6210 }, { "grad_norm": 0.6407209634780884, "learning_rate": 9.978980486956899e-05, "loss": 0.0731, "step": 6220 }, { "grad_norm": 0.4536982774734497, "learning_rate": 9.978790744106332e-05, "loss": 0.0788, "step": 6230 }, { "grad_norm": 0.6783651113510132, "learning_rate": 9.978600150516594e-05, "loss": 0.0827, "step": 6240 }, { "grad_norm": 0.5863257050514221, "learning_rate": 9.978408706220259e-05, "loss": 0.0877, "step": 6250 }, { "grad_norm": 0.6336138844490051, "learning_rate": 9.978216411250032e-05, "loss": 0.066, "step": 6260 }, { "grad_norm": 0.5648318529129028, "learning_rate": 9.978023265638778e-05, "loss": 0.0637, "step": 6270 }, { "grad_norm": 0.48624664545059204, "learning_rate": 9.977829269419495e-05, "loss": 0.0812, "step": 6280 }, { "grad_norm": 0.5213446617126465, "learning_rate": 9.977634422625335e-05, "loss": 0.08, "step": 6290 }, { "grad_norm": 0.42203330993652344, "learning_rate": 9.97743872528959e-05, "loss": 0.0789, "step": 6300 }, { "grad_norm": 0.5439860820770264, "learning_rate": 9.9772421774457e-05, "loss": 0.0679, "step": 6310 }, { "grad_norm": 0.46393829584121704, "learning_rate": 9.977044779127252e-05, "loss": 0.0647, "step": 6320 }, { "grad_norm": 0.6349640488624573, "learning_rate": 9.976846530367971e-05, "loss": 0.0826, "step": 6330 }, { "grad_norm": 0.605890154838562, "learning_rate": 9.976647431201735e-05, "loss": 0.0918, "step": 6340 }, { "grad_norm": 0.5013226866722107, "learning_rate": 9.976447481662568e-05, "loss": 0.0907, "step": 6350 }, { "grad_norm": 0.7366779446601868, "learning_rate": 9.976246681784629e-05, "loss": 0.0882, "step": 6360 }, { "grad_norm": 0.47803711891174316, "learning_rate": 9.976045031602234e-05, "loss": 0.0756, "step": 6370 }, { "grad_norm": 0.7826265096664429, "learning_rate": 9.975842531149837e-05, "loss": 0.069, "step": 6380 }, { "grad_norm": 0.7123594284057617, "learning_rate": 9.975639180462043e-05, "loss": 0.0783, "step": 6390 }, { "grad_norm": 0.569442868232727, "learning_rate": 9.975434979573596e-05, "loss": 0.0736, "step": 6400 }, { "grad_norm": 0.4911363422870636, "learning_rate": 9.97522992851939e-05, "loss": 0.0691, "step": 6410 }, { "grad_norm": 0.42827486991882324, "learning_rate": 9.975024027334461e-05, "loss": 0.0776, "step": 6420 }, { "grad_norm": 0.4566057026386261, "learning_rate": 9.974817276053993e-05, "loss": 0.0679, "step": 6430 }, { "grad_norm": 0.5236285328865051, "learning_rate": 9.974609674713315e-05, "loss": 0.0711, "step": 6440 }, { "grad_norm": 0.5745396614074707, "learning_rate": 9.9744012233479e-05, "loss": 0.0734, "step": 6450 }, { "grad_norm": 0.6976041197776794, "learning_rate": 9.974191921993366e-05, "loss": 0.0788, "step": 6460 }, { "grad_norm": 0.5696637630462646, "learning_rate": 9.973981770685474e-05, "loss": 0.0934, "step": 6470 }, { "grad_norm": 0.5360947251319885, "learning_rate": 9.97377076946014e-05, "loss": 0.0712, "step": 6480 }, { "grad_norm": 0.753373384475708, "learning_rate": 9.973558918353412e-05, "loss": 0.0908, "step": 6490 }, { "grad_norm": 0.3954641819000244, "learning_rate": 9.973346217401494e-05, "loss": 0.0714, "step": 6500 }, { "grad_norm": 0.4446674585342407, "learning_rate": 9.973132666640726e-05, "loss": 0.0662, "step": 6510 }, { "grad_norm": 0.47010567784309387, "learning_rate": 9.972918266107602e-05, "loss": 0.0747, "step": 6520 }, { "grad_norm": 0.43205779790878296, "learning_rate": 9.972703015838756e-05, "loss": 0.0639, "step": 6530 }, { "grad_norm": 0.43647101521492004, "learning_rate": 9.97248691587097e-05, "loss": 0.0748, "step": 6540 }, { "grad_norm": 0.8444690704345703, "learning_rate": 9.972269966241166e-05, "loss": 0.0665, "step": 6550 }, { "grad_norm": 0.4865642488002777, "learning_rate": 9.972052166986417e-05, "loss": 0.0758, "step": 6560 }, { "grad_norm": 0.4752182960510254, "learning_rate": 9.971833518143938e-05, "loss": 0.0886, "step": 6570 }, { "grad_norm": 0.5680209398269653, "learning_rate": 9.971614019751093e-05, "loss": 0.0593, "step": 6580 }, { "grad_norm": 0.5261242985725403, "learning_rate": 9.971393671845383e-05, "loss": 0.0768, "step": 6590 }, { "grad_norm": 0.6020322442054749, "learning_rate": 9.971172474464464e-05, "loss": 0.092, "step": 6600 }, { "grad_norm": 0.4396893084049225, "learning_rate": 9.97095042764613e-05, "loss": 0.0666, "step": 6610 }, { "grad_norm": 0.4705216884613037, "learning_rate": 9.970727531428324e-05, "loss": 0.0641, "step": 6620 }, { "grad_norm": 0.5015608072280884, "learning_rate": 9.970503785849132e-05, "loss": 0.0702, "step": 6630 }, { "grad_norm": 0.4803021252155304, "learning_rate": 9.970279190946788e-05, "loss": 0.0724, "step": 6640 }, { "grad_norm": 0.6008336544036865, "learning_rate": 9.970053746759667e-05, "loss": 0.0878, "step": 6650 }, { "grad_norm": 0.40992945432662964, "learning_rate": 9.969827453326292e-05, "loss": 0.0685, "step": 6660 }, { "grad_norm": 0.6560814380645752, "learning_rate": 9.969600310685332e-05, "loss": 0.0648, "step": 6670 }, { "grad_norm": 0.5757705569267273, "learning_rate": 9.969372318875596e-05, "loss": 0.0769, "step": 6680 }, { "grad_norm": 0.5247528553009033, "learning_rate": 9.969143477936043e-05, "loss": 0.0701, "step": 6690 }, { "grad_norm": 0.4583686292171478, "learning_rate": 9.968913787905775e-05, "loss": 0.0834, "step": 6700 }, { "grad_norm": 0.5549739003181458, "learning_rate": 9.968683248824045e-05, "loss": 0.0922, "step": 6710 }, { "grad_norm": 0.5429378747940063, "learning_rate": 9.968451860730238e-05, "loss": 0.0796, "step": 6720 }, { "grad_norm": 0.5327332615852356, "learning_rate": 9.968219623663896e-05, "loss": 0.0736, "step": 6730 }, { "grad_norm": 0.6038759350776672, "learning_rate": 9.967986537664702e-05, "loss": 0.0939, "step": 6740 }, { "grad_norm": 0.553641140460968, "learning_rate": 9.967752602772483e-05, "loss": 0.0711, "step": 6750 }, { "grad_norm": 0.39873144030570984, "learning_rate": 9.967517819027212e-05, "loss": 0.0787, "step": 6760 }, { "grad_norm": 0.5509161949157715, "learning_rate": 9.967282186469009e-05, "loss": 0.0699, "step": 6770 }, { "grad_norm": 0.5609347820281982, "learning_rate": 9.967045705138135e-05, "loss": 0.0773, "step": 6780 }, { "grad_norm": 0.4277769923210144, "learning_rate": 9.966808375074998e-05, "loss": 0.0665, "step": 6790 }, { "grad_norm": 0.5986961126327515, "learning_rate": 9.966570196320154e-05, "loss": 0.0764, "step": 6800 }, { "grad_norm": 0.6862103939056396, "learning_rate": 9.966331168914299e-05, "loss": 0.073, "step": 6810 }, { "grad_norm": 0.5975576639175415, "learning_rate": 9.966091292898277e-05, "loss": 0.0883, "step": 6820 }, { "grad_norm": 0.5861226916313171, "learning_rate": 9.965850568313076e-05, "loss": 0.073, "step": 6830 }, { "grad_norm": 0.48472860455513, "learning_rate": 9.965608995199827e-05, "loss": 0.0703, "step": 6840 }, { "grad_norm": 0.47803083062171936, "learning_rate": 9.965366573599812e-05, "loss": 0.0704, "step": 6850 }, { "grad_norm": 0.5243198871612549, "learning_rate": 9.965123303554453e-05, "loss": 0.0643, "step": 6860 }, { "grad_norm": 0.7425332069396973, "learning_rate": 9.964879185105317e-05, "loss": 0.0883, "step": 6870 }, { "grad_norm": 0.56512850522995, "learning_rate": 9.964634218294119e-05, "loss": 0.0781, "step": 6880 }, { "grad_norm": 0.44926273822784424, "learning_rate": 9.964388403162714e-05, "loss": 0.0736, "step": 6890 }, { "grad_norm": 0.6101380586624146, "learning_rate": 9.96414173975311e-05, "loss": 0.0791, "step": 6900 }, { "grad_norm": 0.6340954303741455, "learning_rate": 9.963894228107451e-05, "loss": 0.0684, "step": 6910 }, { "grad_norm": 0.42253759503364563, "learning_rate": 9.963645868268032e-05, "loss": 0.0742, "step": 6920 }, { "grad_norm": 0.6031054258346558, "learning_rate": 9.963396660277289e-05, "loss": 0.0761, "step": 6930 }, { "grad_norm": 0.5525968074798584, "learning_rate": 9.963146604177807e-05, "loss": 0.0649, "step": 6940 }, { "grad_norm": 0.40889036655426025, "learning_rate": 9.962895700012311e-05, "loss": 0.0671, "step": 6950 }, { "grad_norm": 0.49907228350639343, "learning_rate": 9.962643947823677e-05, "loss": 0.0732, "step": 6960 }, { "grad_norm": 0.3878808915615082, "learning_rate": 9.962391347654921e-05, "loss": 0.0767, "step": 6970 }, { "grad_norm": 0.4241706430912018, "learning_rate": 9.962137899549204e-05, "loss": 0.0742, "step": 6980 }, { "grad_norm": 0.5566214323043823, "learning_rate": 9.961883603549835e-05, "loss": 0.0587, "step": 6990 }, { "grad_norm": 0.5719122290611267, "learning_rate": 9.961628459700267e-05, "loss": 0.0679, "step": 7000 }, { "grad_norm": 0.42736101150512695, "learning_rate": 9.961372468044095e-05, "loss": 0.0785, "step": 7010 }, { "grad_norm": 0.4645402133464813, "learning_rate": 9.961115628625062e-05, "loss": 0.0662, "step": 7020 }, { "grad_norm": 0.5883333683013916, "learning_rate": 9.960857941487056e-05, "loss": 0.0813, "step": 7030 }, { "grad_norm": 0.5910856127738953, "learning_rate": 9.960599406674106e-05, "loss": 0.0793, "step": 7040 }, { "grad_norm": 0.5083008408546448, "learning_rate": 9.960340024230393e-05, "loss": 0.0653, "step": 7050 }, { "grad_norm": 0.5178506970405579, "learning_rate": 9.960079794200232e-05, "loss": 0.0598, "step": 7060 }, { "grad_norm": 0.5098772644996643, "learning_rate": 9.959818716628096e-05, "loss": 0.0648, "step": 7070 }, { "grad_norm": 0.5778083205223083, "learning_rate": 9.95955679155859e-05, "loss": 0.0777, "step": 7080 }, { "grad_norm": 0.7266832590103149, "learning_rate": 9.959294019036472e-05, "loss": 0.0756, "step": 7090 }, { "grad_norm": 0.5586684346199036, "learning_rate": 9.959030399106646e-05, "loss": 0.0575, "step": 7100 }, { "grad_norm": 0.42340102791786194, "learning_rate": 9.958765931814153e-05, "loss": 0.0804, "step": 7110 }, { "grad_norm": 0.4269759953022003, "learning_rate": 9.958500617204184e-05, "loss": 0.0611, "step": 7120 }, { "grad_norm": 0.40224990248680115, "learning_rate": 9.958234455322075e-05, "loss": 0.0645, "step": 7130 }, { "grad_norm": 0.5037882924079895, "learning_rate": 9.957967446213308e-05, "loss": 0.0779, "step": 7140 }, { "grad_norm": 0.4262818694114685, "learning_rate": 9.957699589923501e-05, "loss": 0.0805, "step": 7150 }, { "grad_norm": 0.6871228814125061, "learning_rate": 9.957430886498431e-05, "loss": 0.0765, "step": 7160 }, { "grad_norm": 0.47919899225234985, "learning_rate": 9.957161335984008e-05, "loss": 0.0725, "step": 7170 }, { "grad_norm": 0.512951672077179, "learning_rate": 9.956890938426291e-05, "loss": 0.0697, "step": 7180 }, { "grad_norm": 0.552742063999176, "learning_rate": 9.956619693871482e-05, "loss": 0.082, "step": 7190 }, { "grad_norm": 0.6340886950492859, "learning_rate": 9.956347602365934e-05, "loss": 0.0781, "step": 7200 }, { "grad_norm": 0.3722259998321533, "learning_rate": 9.956074663956135e-05, "loss": 0.0701, "step": 7210 }, { "grad_norm": 0.4659010171890259, "learning_rate": 9.955800878688726e-05, "loss": 0.0708, "step": 7220 }, { "grad_norm": 0.562798023223877, "learning_rate": 9.955526246610489e-05, "loss": 0.0582, "step": 7230 }, { "grad_norm": 0.6508080363273621, "learning_rate": 9.955250767768349e-05, "loss": 0.0791, "step": 7240 }, { "grad_norm": 0.5090250372886658, "learning_rate": 9.95497444220938e-05, "loss": 0.0786, "step": 7250 }, { "grad_norm": 0.5562170147895813, "learning_rate": 9.954697269980797e-05, "loss": 0.0621, "step": 7260 }, { "grad_norm": 0.42775920033454895, "learning_rate": 9.954419251129962e-05, "loss": 0.0686, "step": 7270 }, { "grad_norm": 0.4655504822731018, "learning_rate": 9.95414038570438e-05, "loss": 0.0643, "step": 7280 }, { "grad_norm": 0.5739526152610779, "learning_rate": 9.953860673751703e-05, "loss": 0.0665, "step": 7290 }, { "grad_norm": 0.40633824467658997, "learning_rate": 9.953580115319725e-05, "loss": 0.0646, "step": 7300 }, { "grad_norm": 0.45352426171302795, "learning_rate": 9.953298710456387e-05, "loss": 0.0695, "step": 7310 }, { "grad_norm": 0.49995431303977966, "learning_rate": 9.953016459209771e-05, "loss": 0.0615, "step": 7320 }, { "grad_norm": 0.4470403492450714, "learning_rate": 9.952733361628108e-05, "loss": 0.0624, "step": 7330 }, { "grad_norm": 0.6651525497436523, "learning_rate": 9.952449417759772e-05, "loss": 0.0663, "step": 7340 }, { "grad_norm": 0.5817423462867737, "learning_rate": 9.952164627653279e-05, "loss": 0.0596, "step": 7350 }, { "grad_norm": 0.6275141835212708, "learning_rate": 9.951878991357292e-05, "loss": 0.0701, "step": 7360 }, { "grad_norm": 0.6088101863861084, "learning_rate": 9.951592508920622e-05, "loss": 0.0856, "step": 7370 }, { "grad_norm": 0.6470544338226318, "learning_rate": 9.951305180392219e-05, "loss": 0.0732, "step": 7380 }, { "grad_norm": 0.6911062002182007, "learning_rate": 9.951017005821178e-05, "loss": 0.0836, "step": 7390 }, { "grad_norm": 0.5548333525657654, "learning_rate": 9.95072798525674e-05, "loss": 0.0666, "step": 7400 }, { "grad_norm": 0.7085157632827759, "learning_rate": 9.950438118748293e-05, "loss": 0.0687, "step": 7410 }, { "grad_norm": 0.6664218306541443, "learning_rate": 9.950147406345366e-05, "loss": 0.0841, "step": 7420 }, { "grad_norm": 0.5424032211303711, "learning_rate": 9.949855848097635e-05, "loss": 0.081, "step": 7430 }, { "grad_norm": 0.5907298922538757, "learning_rate": 9.949563444054916e-05, "loss": 0.0762, "step": 7440 }, { "grad_norm": 0.568148672580719, "learning_rate": 9.949270194267178e-05, "loss": 0.0745, "step": 7450 }, { "grad_norm": 0.38668832182884216, "learning_rate": 9.948976098784526e-05, "loss": 0.0733, "step": 7460 }, { "grad_norm": 0.6174092292785645, "learning_rate": 9.948681157657213e-05, "loss": 0.0825, "step": 7470 }, { "grad_norm": 0.6494572758674622, "learning_rate": 9.948385370935638e-05, "loss": 0.073, "step": 7480 }, { "grad_norm": 0.6952996253967285, "learning_rate": 9.94808873867034e-05, "loss": 0.0821, "step": 7490 }, { "grad_norm": 0.46093201637268066, "learning_rate": 9.947791260912009e-05, "loss": 0.0724, "step": 7500 }, { "grad_norm": 0.638147234916687, "learning_rate": 9.947492937711474e-05, "loss": 0.0598, "step": 7510 }, { "grad_norm": 0.6669626235961914, "learning_rate": 9.947193769119707e-05, "loss": 0.0658, "step": 7520 }, { "grad_norm": 0.43258774280548096, "learning_rate": 9.946893755187834e-05, "loss": 0.0669, "step": 7530 }, { "grad_norm": 0.4724234640598297, "learning_rate": 9.946592895967115e-05, "loss": 0.0716, "step": 7540 }, { "grad_norm": 0.6285163760185242, "learning_rate": 9.94629119150896e-05, "loss": 0.067, "step": 7550 }, { "grad_norm": 0.5335878729820251, "learning_rate": 9.94598864186492e-05, "loss": 0.0692, "step": 7560 }, { "grad_norm": 0.637864351272583, "learning_rate": 9.945685247086696e-05, "loss": 0.0872, "step": 7570 }, { "grad_norm": 0.5769896507263184, "learning_rate": 9.945381007226129e-05, "loss": 0.0736, "step": 7580 }, { "grad_norm": 0.5209776759147644, "learning_rate": 9.945075922335203e-05, "loss": 0.0866, "step": 7590 }, { "grad_norm": 0.6545878648757935, "learning_rate": 9.944769992466049e-05, "loss": 0.0945, "step": 7600 }, { "grad_norm": 0.7306598424911499, "learning_rate": 9.944463217670945e-05, "loss": 0.0853, "step": 7610 }, { "grad_norm": 0.7218620181083679, "learning_rate": 9.944155598002307e-05, "loss": 0.0843, "step": 7620 }, { "grad_norm": 0.5782116055488586, "learning_rate": 9.943847133512701e-05, "loss": 0.0826, "step": 7630 }, { "grad_norm": 0.6278444528579712, "learning_rate": 9.943537824254834e-05, "loss": 0.0694, "step": 7640 }, { "grad_norm": 0.45167598128318787, "learning_rate": 9.943227670281559e-05, "loss": 0.0683, "step": 7650 }, { "grad_norm": 0.4294147193431854, "learning_rate": 9.942916671645873e-05, "loss": 0.066, "step": 7660 }, { "grad_norm": 0.5129007697105408, "learning_rate": 9.942604828400916e-05, "loss": 0.0726, "step": 7670 }, { "grad_norm": 0.5088673830032349, "learning_rate": 9.942292140599975e-05, "loss": 0.0785, "step": 7680 }, { "grad_norm": 0.5110358595848083, "learning_rate": 9.94197860829648e-05, "loss": 0.072, "step": 7690 }, { "grad_norm": 0.6397515535354614, "learning_rate": 9.941664231544004e-05, "loss": 0.0619, "step": 7700 }, { "grad_norm": 0.5875016450881958, "learning_rate": 9.941349010396264e-05, "loss": 0.077, "step": 7710 }, { "grad_norm": 0.5534569025039673, "learning_rate": 9.941032944907125e-05, "loss": 0.0703, "step": 7720 }, { "grad_norm": 0.6002580523490906, "learning_rate": 9.940716035130596e-05, "loss": 0.0718, "step": 7730 }, { "grad_norm": 0.5955712199211121, "learning_rate": 9.940398281120821e-05, "loss": 0.0748, "step": 7740 }, { "grad_norm": 0.5791760087013245, "learning_rate": 9.940079682932102e-05, "loss": 0.0635, "step": 7750 }, { "grad_norm": 0.4606814384460449, "learning_rate": 9.939760240618877e-05, "loss": 0.0749, "step": 7760 }, { "grad_norm": 0.48981809616088867, "learning_rate": 9.939439954235729e-05, "loss": 0.0829, "step": 7770 }, { "grad_norm": 0.5540587902069092, "learning_rate": 9.939118823837387e-05, "loss": 0.087, "step": 7780 }, { "grad_norm": 0.5771268010139465, "learning_rate": 9.938796849478725e-05, "loss": 0.0773, "step": 7790 }, { "grad_norm": 0.4915456771850586, "learning_rate": 9.938474031214755e-05, "loss": 0.0742, "step": 7800 }, { "grad_norm": 0.5182307362556458, "learning_rate": 9.938150369100643e-05, "loss": 0.0777, "step": 7810 }, { "grad_norm": 0.41860663890838623, "learning_rate": 9.93782586319169e-05, "loss": 0.0588, "step": 7820 }, { "grad_norm": 0.605725884437561, "learning_rate": 9.937500513543348e-05, "loss": 0.0766, "step": 7830 }, { "grad_norm": 0.6915348768234253, "learning_rate": 9.937174320211207e-05, "loss": 0.0628, "step": 7840 }, { "grad_norm": 0.37630435824394226, "learning_rate": 9.936847283251009e-05, "loss": 0.0681, "step": 7850 }, { "grad_norm": 0.567507266998291, "learning_rate": 9.936519402718632e-05, "loss": 0.0797, "step": 7860 }, { "grad_norm": 0.5312703251838684, "learning_rate": 9.936190678670102e-05, "loss": 0.0831, "step": 7870 }, { "grad_norm": 0.5668172836303711, "learning_rate": 9.935861111161593e-05, "loss": 0.0755, "step": 7880 }, { "grad_norm": 0.5476062297821045, "learning_rate": 9.935530700249416e-05, "loss": 0.0705, "step": 7890 }, { "grad_norm": 0.5177210569381714, "learning_rate": 9.935199445990028e-05, "loss": 0.082, "step": 7900 }, { "grad_norm": 0.4673998951911926, "learning_rate": 9.934867348440033e-05, "loss": 0.0754, "step": 7910 }, { "grad_norm": 0.49662935733795166, "learning_rate": 9.934534407656176e-05, "loss": 0.0948, "step": 7920 }, { "grad_norm": 0.5975887775421143, "learning_rate": 9.93420062369535e-05, "loss": 0.0724, "step": 7930 }, { "grad_norm": 0.7228508591651917, "learning_rate": 9.933865996614589e-05, "loss": 0.0655, "step": 7940 }, { "grad_norm": 0.47130468487739563, "learning_rate": 9.933530526471068e-05, "loss": 0.0676, "step": 7950 }, { "grad_norm": 0.6960284113883972, "learning_rate": 9.933194213322114e-05, "loss": 0.0707, "step": 7960 }, { "grad_norm": 0.5211763978004456, "learning_rate": 9.932857057225192e-05, "loss": 0.0655, "step": 7970 }, { "grad_norm": 0.5461847186088562, "learning_rate": 9.932519058237912e-05, "loss": 0.066, "step": 7980 }, { "grad_norm": 0.5489415526390076, "learning_rate": 9.932180216418032e-05, "loss": 0.074, "step": 7990 }, { "grad_norm": 0.5305334329605103, "learning_rate": 9.931840531823446e-05, "loss": 0.0684, "step": 8000 }, { "grad_norm": 0.5026463866233826, "learning_rate": 9.9315000045122e-05, "loss": 0.055, "step": 8010 }, { "grad_norm": 0.46939805150032043, "learning_rate": 9.931158634542481e-05, "loss": 0.0583, "step": 8020 }, { "grad_norm": 0.400319904088974, "learning_rate": 9.930816421972617e-05, "loss": 0.057, "step": 8030 }, { "grad_norm": 0.5571240186691284, "learning_rate": 9.930473366861086e-05, "loss": 0.0629, "step": 8040 }, { "grad_norm": 0.5526134371757507, "learning_rate": 9.930129469266505e-05, "loss": 0.0785, "step": 8050 }, { "grad_norm": 0.4649295210838318, "learning_rate": 9.929784729247638e-05, "loss": 0.0556, "step": 8060 }, { "grad_norm": 0.4749349355697632, "learning_rate": 9.929439146863389e-05, "loss": 0.068, "step": 8070 }, { "grad_norm": 0.7651559710502625, "learning_rate": 9.92909272217281e-05, "loss": 0.0867, "step": 8080 }, { "grad_norm": 0.588773787021637, "learning_rate": 9.928745455235097e-05, "loss": 0.0651, "step": 8090 }, { "grad_norm": 0.46984565258026123, "learning_rate": 9.928397346109588e-05, "loss": 0.073, "step": 8100 }, { "grad_norm": 0.6393337249755859, "learning_rate": 9.928048394855762e-05, "loss": 0.0718, "step": 8110 }, { "grad_norm": 0.6002474427223206, "learning_rate": 9.92769860153325e-05, "loss": 0.0755, "step": 8120 }, { "grad_norm": 0.5276062488555908, "learning_rate": 9.927347966201819e-05, "loss": 0.0839, "step": 8130 }, { "grad_norm": 0.5478351712226868, "learning_rate": 9.926996488921383e-05, "loss": 0.0728, "step": 8140 }, { "grad_norm": 0.7426574230194092, "learning_rate": 9.926644169752001e-05, "loss": 0.0649, "step": 8150 }, { "grad_norm": 0.5403677821159363, "learning_rate": 9.926291008753875e-05, "loss": 0.0821, "step": 8160 }, { "grad_norm": 0.397549033164978, "learning_rate": 9.92593700598735e-05, "loss": 0.0749, "step": 8170 }, { "grad_norm": 0.47741299867630005, "learning_rate": 9.925582161512915e-05, "loss": 0.0646, "step": 8180 }, { "grad_norm": 0.48344624042510986, "learning_rate": 9.925226475391205e-05, "loss": 0.0702, "step": 8190 }, { "grad_norm": 0.5764510631561279, "learning_rate": 9.924869947682993e-05, "loss": 0.0815, "step": 8200 }, { "grad_norm": 0.6187590956687927, "learning_rate": 9.924512578449204e-05, "loss": 0.0698, "step": 8210 }, { "grad_norm": 0.5715123414993286, "learning_rate": 9.924154367750901e-05, "loss": 0.0801, "step": 8220 }, { "grad_norm": 0.7509397268295288, "learning_rate": 9.923795315649293e-05, "loss": 0.0732, "step": 8230 }, { "grad_norm": 0.36518344283103943, "learning_rate": 9.92343542220573e-05, "loss": 0.0713, "step": 8240 }, { "grad_norm": 0.5509667992591858, "learning_rate": 9.92307468748171e-05, "loss": 0.0665, "step": 8250 }, { "grad_norm": 0.4835318624973297, "learning_rate": 9.922713111538873e-05, "loss": 0.0544, "step": 8260 }, { "grad_norm": 0.5934747457504272, "learning_rate": 9.922350694439003e-05, "loss": 0.0648, "step": 8270 }, { "grad_norm": 0.6826234459877014, "learning_rate": 9.921987436244024e-05, "loss": 0.0767, "step": 8280 }, { "grad_norm": 0.5561403036117554, "learning_rate": 9.921623337016008e-05, "loss": 0.06, "step": 8290 }, { "grad_norm": 0.38178348541259766, "learning_rate": 9.921258396817172e-05, "loss": 0.0693, "step": 8300 }, { "grad_norm": 0.4130634069442749, "learning_rate": 9.920892615709874e-05, "loss": 0.0635, "step": 8310 }, { "grad_norm": 0.4847143590450287, "learning_rate": 9.920525993756612e-05, "loss": 0.0572, "step": 8320 }, { "grad_norm": 0.5477247834205627, "learning_rate": 9.920158531020036e-05, "loss": 0.0707, "step": 8330 }, { "grad_norm": 0.4392712712287903, "learning_rate": 9.919790227562933e-05, "loss": 0.0576, "step": 8340 }, { "grad_norm": 0.446607381105423, "learning_rate": 9.919421083448237e-05, "loss": 0.075, "step": 8350 }, { "grad_norm": 0.4804190695285797, "learning_rate": 9.919051098739022e-05, "loss": 0.057, "step": 8360 }, { "grad_norm": 0.6296531558036804, "learning_rate": 9.918680273498514e-05, "loss": 0.0602, "step": 8370 }, { "grad_norm": 0.5327516794204712, "learning_rate": 9.918308607790072e-05, "loss": 0.0724, "step": 8380 }, { "grad_norm": 0.34504905343055725, "learning_rate": 9.917936101677205e-05, "loss": 0.0582, "step": 8390 }, { "grad_norm": 0.476820170879364, "learning_rate": 9.917562755223564e-05, "loss": 0.0627, "step": 8400 }, { "grad_norm": 0.5608691573143005, "learning_rate": 9.917188568492944e-05, "loss": 0.072, "step": 8410 }, { "grad_norm": 0.6600950956344604, "learning_rate": 9.916813541549283e-05, "loss": 0.0612, "step": 8420 }, { "grad_norm": 0.5145482420921326, "learning_rate": 9.916437674456663e-05, "loss": 0.061, "step": 8430 }, { "grad_norm": 0.5666571855545044, "learning_rate": 9.916060967279308e-05, "loss": 0.0639, "step": 8440 }, { "grad_norm": 0.8221482038497925, "learning_rate": 9.91568342008159e-05, "loss": 0.0598, "step": 8450 }, { "grad_norm": 0.4483817517757416, "learning_rate": 9.915305032928019e-05, "loss": 0.0706, "step": 8460 }, { "grad_norm": 0.5115541815757751, "learning_rate": 9.914925805883253e-05, "loss": 0.0681, "step": 8470 }, { "grad_norm": 0.6148876547813416, "learning_rate": 9.914545739012088e-05, "loss": 0.0593, "step": 8480 }, { "grad_norm": 0.6512833833694458, "learning_rate": 9.91416483237947e-05, "loss": 0.0712, "step": 8490 }, { "grad_norm": 0.5601624250411987, "learning_rate": 9.913783086050485e-05, "loss": 0.0852, "step": 8500 }, { "grad_norm": 0.46689626574516296, "learning_rate": 9.913400500090364e-05, "loss": 0.0713, "step": 8510 }, { "grad_norm": 0.4708261489868164, "learning_rate": 9.913017074564479e-05, "loss": 0.057, "step": 8520 }, { "grad_norm": 0.49818921089172363, "learning_rate": 9.912632809538348e-05, "loss": 0.0668, "step": 8530 }, { "grad_norm": 0.5648747086524963, "learning_rate": 9.912247705077629e-05, "loss": 0.0707, "step": 8540 }, { "grad_norm": 0.3665335178375244, "learning_rate": 9.911861761248127e-05, "loss": 0.0739, "step": 8550 }, { "grad_norm": 0.5577801465988159, "learning_rate": 9.91147497811579e-05, "loss": 0.0699, "step": 8560 }, { "grad_norm": 0.6129829287528992, "learning_rate": 9.911087355746709e-05, "loss": 0.0992, "step": 8570 }, { "grad_norm": 0.5776244401931763, "learning_rate": 9.910698894207117e-05, "loss": 0.0758, "step": 8580 }, { "grad_norm": 0.3401185870170593, "learning_rate": 9.910309593563392e-05, "loss": 0.0672, "step": 8590 }, { "grad_norm": 0.5741989612579346, "learning_rate": 9.909919453882057e-05, "loss": 0.0572, "step": 8600 }, { "grad_norm": 0.6111934185028076, "learning_rate": 9.90952847522977e-05, "loss": 0.0587, "step": 8610 }, { "grad_norm": 0.658676266670227, "learning_rate": 9.909136657673346e-05, "loss": 0.0691, "step": 8620 }, { "grad_norm": 0.36490577459335327, "learning_rate": 9.908744001279731e-05, "loss": 0.0532, "step": 8630 }, { "grad_norm": 0.7059012055397034, "learning_rate": 9.90835050611602e-05, "loss": 0.068, "step": 8640 }, { "grad_norm": 0.5558072328567505, "learning_rate": 9.90795617224945e-05, "loss": 0.0694, "step": 8650 }, { "grad_norm": 0.5859278440475464, "learning_rate": 9.907560999747405e-05, "loss": 0.0747, "step": 8660 }, { "grad_norm": 0.6196101307868958, "learning_rate": 9.907164988677408e-05, "loss": 0.0618, "step": 8670 }, { "grad_norm": 0.4346092641353607, "learning_rate": 9.906768139107124e-05, "loss": 0.0551, "step": 8680 }, { "grad_norm": 0.5611621141433716, "learning_rate": 9.906370451104367e-05, "loss": 0.0599, "step": 8690 }, { "grad_norm": 0.3849419355392456, "learning_rate": 9.905971924737088e-05, "loss": 0.0521, "step": 8700 }, { "grad_norm": 0.46514609456062317, "learning_rate": 9.905572560073387e-05, "loss": 0.0607, "step": 8710 }, { "grad_norm": 1.2180532217025757, "learning_rate": 9.905172357181501e-05, "loss": 0.078, "step": 8720 }, { "grad_norm": 0.6304105520248413, "learning_rate": 9.904771316129817e-05, "loss": 0.0627, "step": 8730 }, { "grad_norm": 0.514057993888855, "learning_rate": 9.904369436986862e-05, "loss": 0.0561, "step": 8740 }, { "grad_norm": 0.4590342938899994, "learning_rate": 9.903966719821303e-05, "loss": 0.0708, "step": 8750 }, { "grad_norm": 0.3894416093826294, "learning_rate": 9.903563164701956e-05, "loss": 0.0591, "step": 8760 }, { "grad_norm": 0.5768414735794067, "learning_rate": 9.903158771697778e-05, "loss": 0.065, "step": 8770 }, { "grad_norm": 0.6638126969337463, "learning_rate": 9.902753540877867e-05, "loss": 0.0863, "step": 8780 }, { "grad_norm": 0.6287906169891357, "learning_rate": 9.902347472311466e-05, "loss": 0.0612, "step": 8790 }, { "grad_norm": 0.530501663684845, "learning_rate": 9.901940566067962e-05, "loss": 0.0471, "step": 8800 }, { "grad_norm": 0.7911847829818726, "learning_rate": 9.901532822216883e-05, "loss": 0.0631, "step": 8810 }, { "grad_norm": 0.4793124794960022, "learning_rate": 9.901124240827904e-05, "loss": 0.0599, "step": 8820 }, { "grad_norm": 0.6220221519470215, "learning_rate": 9.900714821970835e-05, "loss": 0.0645, "step": 8830 }, { "grad_norm": 0.5031537413597107, "learning_rate": 9.900304565715641e-05, "loss": 0.0554, "step": 8840 }, { "grad_norm": 0.40934282541275024, "learning_rate": 9.899893472132419e-05, "loss": 0.0554, "step": 8850 }, { "grad_norm": 0.45007774233818054, "learning_rate": 9.899481541291415e-05, "loss": 0.0555, "step": 8860 }, { "grad_norm": 0.6077027916908264, "learning_rate": 9.899068773263016e-05, "loss": 0.0575, "step": 8870 }, { "grad_norm": 0.4991740584373474, "learning_rate": 9.898655168117754e-05, "loss": 0.0679, "step": 8880 }, { "grad_norm": 0.4382721781730652, "learning_rate": 9.898240725926302e-05, "loss": 0.0712, "step": 8890 }, { "grad_norm": 0.5605423450469971, "learning_rate": 9.897825446759478e-05, "loss": 0.0634, "step": 8900 }, { "grad_norm": 0.4959396719932556, "learning_rate": 9.897409330688241e-05, "loss": 0.0656, "step": 8910 }, { "grad_norm": 0.7258632183074951, "learning_rate": 9.896992377783692e-05, "loss": 0.0817, "step": 8920 }, { "grad_norm": 0.5386077165603638, "learning_rate": 9.89657458811708e-05, "loss": 0.0555, "step": 8930 }, { "grad_norm": 0.482919842004776, "learning_rate": 9.896155961759792e-05, "loss": 0.0668, "step": 8940 }, { "grad_norm": 0.3844031095504761, "learning_rate": 9.895736498783361e-05, "loss": 0.06, "step": 8950 }, { "grad_norm": 0.47559934854507446, "learning_rate": 9.895316199259462e-05, "loss": 0.0528, "step": 8960 }, { "grad_norm": 0.5538420677185059, "learning_rate": 9.894895063259909e-05, "loss": 0.0589, "step": 8970 }, { "grad_norm": 0.4515862464904785, "learning_rate": 9.894473090856667e-05, "loss": 0.0703, "step": 8980 }, { "grad_norm": 0.5099793672561646, "learning_rate": 9.894050282121839e-05, "loss": 0.0814, "step": 8990 }, { "grad_norm": 0.5413684844970703, "learning_rate": 9.893626637127668e-05, "loss": 0.0652, "step": 9000 }, { "grad_norm": 0.5396822690963745, "learning_rate": 9.893202155946546e-05, "loss": 0.06, "step": 9010 }, { "grad_norm": 0.49991336464881897, "learning_rate": 9.892776838651006e-05, "loss": 0.0564, "step": 9020 }, { "grad_norm": 0.4001864790916443, "learning_rate": 9.892350685313722e-05, "loss": 0.0712, "step": 9030 }, { "grad_norm": 0.4985629618167877, "learning_rate": 9.891923696007513e-05, "loss": 0.0727, "step": 9040 }, { "grad_norm": 0.518212616443634, "learning_rate": 9.891495870805336e-05, "loss": 0.0587, "step": 9050 }, { "grad_norm": 0.4742549955844879, "learning_rate": 9.891067209780298e-05, "loss": 0.074, "step": 9060 }, { "grad_norm": 0.4788854420185089, "learning_rate": 9.890637713005646e-05, "loss": 0.0582, "step": 9070 }, { "grad_norm": 0.4043080806732178, "learning_rate": 9.890207380554767e-05, "loss": 0.0565, "step": 9080 }, { "grad_norm": 0.45692190527915955, "learning_rate": 9.889776212501196e-05, "loss": 0.0739, "step": 9090 }, { "grad_norm": 0.590097188949585, "learning_rate": 9.889344208918605e-05, "loss": 0.0761, "step": 9100 }, { "grad_norm": 0.51753169298172, "learning_rate": 9.888911369880812e-05, "loss": 0.0729, "step": 9110 }, { "grad_norm": 0.548407793045044, "learning_rate": 9.888477695461777e-05, "loss": 0.0686, "step": 9120 }, { "grad_norm": 0.5873236656188965, "learning_rate": 9.888043185735607e-05, "loss": 0.0743, "step": 9130 }, { "grad_norm": 0.5102695226669312, "learning_rate": 9.887607840776542e-05, "loss": 0.0581, "step": 9140 }, { "grad_norm": 0.5121919512748718, "learning_rate": 9.887171660658975e-05, "loss": 0.0747, "step": 9150 }, { "grad_norm": 0.5045121312141418, "learning_rate": 9.886734645457435e-05, "loss": 0.0647, "step": 9160 }, { "grad_norm": 0.5186971426010132, "learning_rate": 9.886296795246597e-05, "loss": 0.0554, "step": 9170 }, { "grad_norm": 0.4293360114097595, "learning_rate": 9.885858110101276e-05, "loss": 0.057, "step": 9180 }, { "grad_norm": 0.4659256339073181, "learning_rate": 9.885418590096434e-05, "loss": 0.0525, "step": 9190 }, { "grad_norm": 0.4716194272041321, "learning_rate": 9.88497823530717e-05, "loss": 0.0616, "step": 9200 }, { "grad_norm": 0.8094573616981506, "learning_rate": 9.884537045808732e-05, "loss": 0.075, "step": 9210 }, { "grad_norm": 0.646711528301239, "learning_rate": 9.884095021676502e-05, "loss": 0.072, "step": 9220 }, { "grad_norm": 0.5083968639373779, "learning_rate": 9.883652162986017e-05, "loss": 0.0633, "step": 9230 }, { "grad_norm": 0.3991420269012451, "learning_rate": 9.883208469812943e-05, "loss": 0.0659, "step": 9240 }, { "grad_norm": 0.40358710289001465, "learning_rate": 9.882763942233098e-05, "loss": 0.0525, "step": 9250 }, { "grad_norm": 0.337816059589386, "learning_rate": 9.882318580322441e-05, "loss": 0.0473, "step": 9260 }, { "grad_norm": 0.5199457406997681, "learning_rate": 9.881872384157067e-05, "loss": 0.0593, "step": 9270 }, { "grad_norm": 0.5460781455039978, "learning_rate": 9.881425353813225e-05, "loss": 0.0612, "step": 9280 }, { "grad_norm": 0.7871191501617432, "learning_rate": 9.880977489367296e-05, "loss": 0.0779, "step": 9290 }, { "grad_norm": 0.5775189995765686, "learning_rate": 9.88052879089581e-05, "loss": 0.0671, "step": 9300 }, { "grad_norm": 0.5806031823158264, "learning_rate": 9.880079258475434e-05, "loss": 0.0651, "step": 9310 }, { "grad_norm": 0.4353421926498413, "learning_rate": 9.879628892182985e-05, "loss": 0.0839, "step": 9320 }, { "grad_norm": 0.4343526065349579, "learning_rate": 9.879177692095416e-05, "loss": 0.0541, "step": 9330 }, { "grad_norm": 0.3865508735179901, "learning_rate": 9.878725658289825e-05, "loss": 0.0653, "step": 9340 }, { "grad_norm": 0.4435287415981293, "learning_rate": 9.878272790843454e-05, "loss": 0.0594, "step": 9350 }, { "grad_norm": 0.49283698201179504, "learning_rate": 9.877819089833682e-05, "loss": 0.0617, "step": 9360 }, { "grad_norm": 0.48494160175323486, "learning_rate": 9.877364555338038e-05, "loss": 0.0758, "step": 9370 }, { "grad_norm": 0.4741416573524475, "learning_rate": 9.876909187434186e-05, "loss": 0.0683, "step": 9380 }, { "grad_norm": 0.6143312454223633, "learning_rate": 9.876452986199939e-05, "loss": 0.0649, "step": 9390 }, { "grad_norm": 0.6546781063079834, "learning_rate": 9.875995951713248e-05, "loss": 0.0858, "step": 9400 }, { "grad_norm": 0.5522384643554688, "learning_rate": 9.875538084052207e-05, "loss": 0.0764, "step": 9410 }, { "grad_norm": 0.6614217162132263, "learning_rate": 9.875079383295053e-05, "loss": 0.0631, "step": 9420 }, { "grad_norm": 0.39079734683036804, "learning_rate": 9.874619849520167e-05, "loss": 0.0591, "step": 9430 }, { "grad_norm": 0.5345883369445801, "learning_rate": 9.874159482806069e-05, "loss": 0.0612, "step": 9440 }, { "grad_norm": 0.4843159317970276, "learning_rate": 9.873698283231426e-05, "loss": 0.058, "step": 9450 }, { "grad_norm": 0.5161512494087219, "learning_rate": 9.87323625087504e-05, "loss": 0.0596, "step": 9460 }, { "grad_norm": 0.39223775267601013, "learning_rate": 9.872773385815863e-05, "loss": 0.0691, "step": 9470 }, { "grad_norm": 0.3916289210319519, "learning_rate": 9.872309688132986e-05, "loss": 0.062, "step": 9480 }, { "grad_norm": 0.4784950315952301, "learning_rate": 9.871845157905639e-05, "loss": 0.0562, "step": 9490 }, { "grad_norm": 0.38460296392440796, "learning_rate": 9.871379795213201e-05, "loss": 0.0517, "step": 9500 }, { "grad_norm": 0.4494832158088684, "learning_rate": 9.87091360013519e-05, "loss": 0.0612, "step": 9510 }, { "grad_norm": 0.47170281410217285, "learning_rate": 9.870446572751262e-05, "loss": 0.0567, "step": 9520 }, { "grad_norm": 0.4008587896823883, "learning_rate": 9.869978713141224e-05, "loss": 0.0658, "step": 9530 }, { "grad_norm": 0.8318947553634644, "learning_rate": 9.869510021385016e-05, "loss": 0.0648, "step": 9540 }, { "grad_norm": 0.37689778208732605, "learning_rate": 9.869040497562727e-05, "loss": 0.0554, "step": 9550 }, { "grad_norm": 0.40865856409072876, "learning_rate": 9.868570141754587e-05, "loss": 0.0673, "step": 9560 }, { "grad_norm": 0.4978630542755127, "learning_rate": 9.868098954040965e-05, "loss": 0.0786, "step": 9570 }, { "grad_norm": 0.4341464638710022, "learning_rate": 9.867626934502374e-05, "loss": 0.0733, "step": 9580 }, { "grad_norm": 0.4753769040107727, "learning_rate": 9.86715408321947e-05, "loss": 0.0648, "step": 9590 }, { "grad_norm": 0.4709100127220154, "learning_rate": 9.86668040027305e-05, "loss": 0.0622, "step": 9600 }, { "grad_norm": 0.6122444868087769, "learning_rate": 9.866205885744053e-05, "loss": 0.0576, "step": 9610 }, { "grad_norm": 0.35962623357772827, "learning_rate": 9.865730539713563e-05, "loss": 0.0614, "step": 9620 }, { "grad_norm": 0.3540751039981842, "learning_rate": 9.8652543622628e-05, "loss": 0.0522, "step": 9630 }, { "grad_norm": 0.5111059546470642, "learning_rate": 9.864777353473132e-05, "loss": 0.0594, "step": 9640 }, { "grad_norm": 0.42262983322143555, "learning_rate": 9.864299513426068e-05, "loss": 0.0657, "step": 9650 }, { "grad_norm": 0.7117483615875244, "learning_rate": 9.863820842203254e-05, "loss": 0.0712, "step": 9660 }, { "grad_norm": 0.45196664333343506, "learning_rate": 9.863341339886483e-05, "loss": 0.0629, "step": 9670 }, { "grad_norm": 0.4458892345428467, "learning_rate": 9.86286100655769e-05, "loss": 0.0501, "step": 9680 }, { "grad_norm": 0.5715081691741943, "learning_rate": 9.862379842298953e-05, "loss": 0.0636, "step": 9690 }, { "grad_norm": 0.5373542308807373, "learning_rate": 9.861897847192485e-05, "loss": 0.0526, "step": 9700 }, { "grad_norm": 0.50568026304245, "learning_rate": 9.86141502132065e-05, "loss": 0.0566, "step": 9710 }, { "grad_norm": 0.6585437655448914, "learning_rate": 9.860931364765946e-05, "loss": 0.0626, "step": 9720 }, { "grad_norm": 0.5410544872283936, "learning_rate": 9.860446877611021e-05, "loss": 0.0782, "step": 9730 }, { "grad_norm": 0.6307003498077393, "learning_rate": 9.859961559938655e-05, "loss": 0.0598, "step": 9740 }, { "grad_norm": 0.6375135779380798, "learning_rate": 9.85947541183178e-05, "loss": 0.0588, "step": 9750 }, { "grad_norm": 0.4459591805934906, "learning_rate": 9.858988433373463e-05, "loss": 0.0625, "step": 9760 }, { "grad_norm": 0.5166038274765015, "learning_rate": 9.858500624646918e-05, "loss": 0.0568, "step": 9770 }, { "grad_norm": 0.5329096913337708, "learning_rate": 9.858011985735497e-05, "loss": 0.0603, "step": 9780 }, { "grad_norm": 0.6364641785621643, "learning_rate": 9.857522516722693e-05, "loss": 0.0642, "step": 9790 }, { "grad_norm": 0.4057740569114685, "learning_rate": 9.857032217692145e-05, "loss": 0.057, "step": 9800 }, { "grad_norm": 0.580600380897522, "learning_rate": 9.856541088727631e-05, "loss": 0.0623, "step": 9810 }, { "grad_norm": 0.40823718905448914, "learning_rate": 9.856049129913072e-05, "loss": 0.0669, "step": 9820 }, { "grad_norm": 0.6637150645256042, "learning_rate": 9.85555634133253e-05, "loss": 0.0744, "step": 9830 }, { "grad_norm": 0.440902441740036, "learning_rate": 9.855062723070208e-05, "loss": 0.0868, "step": 9840 }, { "grad_norm": 0.49749889969825745, "learning_rate": 9.854568275210454e-05, "loss": 0.0806, "step": 9850 }, { "grad_norm": 0.47547581791877747, "learning_rate": 9.854072997837754e-05, "loss": 0.0522, "step": 9860 }, { "grad_norm": 0.32597023248672485, "learning_rate": 9.853576891036737e-05, "loss": 0.052, "step": 9870 }, { "grad_norm": 0.4310333728790283, "learning_rate": 9.853079954892177e-05, "loss": 0.0703, "step": 9880 }, { "grad_norm": 1.069735050201416, "learning_rate": 9.852582189488983e-05, "loss": 0.0625, "step": 9890 }, { "grad_norm": 0.5068820714950562, "learning_rate": 9.852083594912212e-05, "loss": 0.0736, "step": 9900 }, { "grad_norm": 0.40432214736938477, "learning_rate": 9.851584171247058e-05, "loss": 0.0577, "step": 9910 }, { "grad_norm": 0.5797638893127441, "learning_rate": 9.851083918578863e-05, "loss": 0.0692, "step": 9920 }, { "grad_norm": 2.3921546936035156, "learning_rate": 9.850582836993103e-05, "loss": 0.0632, "step": 9930 }, { "grad_norm": 0.4243583083152771, "learning_rate": 9.850080926575397e-05, "loss": 0.0583, "step": 9940 }, { "grad_norm": 0.6826049089431763, "learning_rate": 9.849578187411515e-05, "loss": 0.0551, "step": 9950 }, { "grad_norm": 0.7416923642158508, "learning_rate": 9.849074619587354e-05, "loss": 0.0615, "step": 9960 }, { "grad_norm": 0.5708524584770203, "learning_rate": 9.848570223188964e-05, "loss": 0.071, "step": 9970 }, { "grad_norm": 0.44810083508491516, "learning_rate": 9.848064998302531e-05, "loss": 0.0668, "step": 9980 }, { "grad_norm": 0.4796082377433777, "learning_rate": 9.847558945014386e-05, "loss": 0.0643, "step": 9990 }, { "grad_norm": 0.5560008883476257, "learning_rate": 9.847052063410996e-05, "loss": 0.08, "step": 10000 }, { "grad_norm": 0.5190732479095459, "learning_rate": 9.846544353578977e-05, "loss": 0.0735, "step": 10010 }, { "grad_norm": 0.5853862166404724, "learning_rate": 9.846035815605081e-05, "loss": 0.074, "step": 10020 }, { "grad_norm": 0.6423624157905579, "learning_rate": 9.845526449576204e-05, "loss": 0.0635, "step": 10030 }, { "grad_norm": 0.5829293131828308, "learning_rate": 9.845016255579383e-05, "loss": 0.0724, "step": 10040 }, { "grad_norm": 0.46234145760536194, "learning_rate": 9.844505233701794e-05, "loss": 0.0586, "step": 10050 }, { "grad_norm": 0.5466768145561218, "learning_rate": 9.843993384030757e-05, "loss": 0.0586, "step": 10060 }, { "grad_norm": 0.3769165575504303, "learning_rate": 9.843480706653737e-05, "loss": 0.049, "step": 10070 }, { "grad_norm": 0.4533079266548157, "learning_rate": 9.84296720165833e-05, "loss": 0.0597, "step": 10080 }, { "grad_norm": 0.47360390424728394, "learning_rate": 9.842452869132286e-05, "loss": 0.0616, "step": 10090 }, { "grad_norm": 0.4673602283000946, "learning_rate": 9.841937709163489e-05, "loss": 0.0628, "step": 10100 }, { "grad_norm": 0.48849326372146606, "learning_rate": 9.841421721839962e-05, "loss": 0.0489, "step": 10110 }, { "grad_norm": 0.39893975853919983, "learning_rate": 9.840904907249879e-05, "loss": 0.0608, "step": 10120 }, { "grad_norm": 0.6392306089401245, "learning_rate": 9.840387265481545e-05, "loss": 0.0621, "step": 10130 }, { "grad_norm": 0.5340145826339722, "learning_rate": 9.839868796623411e-05, "loss": 0.0579, "step": 10140 }, { "grad_norm": 0.37236496806144714, "learning_rate": 9.839349500764072e-05, "loss": 0.0791, "step": 10150 }, { "grad_norm": 0.542078971862793, "learning_rate": 9.83882937799226e-05, "loss": 0.0699, "step": 10160 }, { "grad_norm": 0.4789299964904785, "learning_rate": 9.838308428396849e-05, "loss": 0.0639, "step": 10170 }, { "grad_norm": 0.4566883444786072, "learning_rate": 9.837786652066854e-05, "loss": 0.074, "step": 10180 }, { "grad_norm": 0.7054589986801147, "learning_rate": 9.837264049091437e-05, "loss": 0.0794, "step": 10190 }, { "grad_norm": 0.8793331980705261, "learning_rate": 9.836740619559893e-05, "loss": 0.0731, "step": 10200 }, { "grad_norm": 0.4720557928085327, "learning_rate": 9.836216363561659e-05, "loss": 0.0683, "step": 10210 }, { "grad_norm": 0.661611795425415, "learning_rate": 9.835691281186322e-05, "loss": 0.0628, "step": 10220 }, { "grad_norm": 0.535715639591217, "learning_rate": 9.8351653725236e-05, "loss": 0.061, "step": 10230 }, { "grad_norm": 0.5547707676887512, "learning_rate": 9.83463863766336e-05, "loss": 0.0817, "step": 10240 }, { "grad_norm": 0.5101153254508972, "learning_rate": 9.834111076695602e-05, "loss": 0.0686, "step": 10250 }, { "grad_norm": 0.3727988302707672, "learning_rate": 9.833582689710477e-05, "loss": 0.0557, "step": 10260 }, { "grad_norm": 0.5597397685050964, "learning_rate": 9.833053476798268e-05, "loss": 0.0602, "step": 10270 }, { "grad_norm": 0.8067002892494202, "learning_rate": 9.832523438049404e-05, "loss": 0.0649, "step": 10280 }, { "grad_norm": 0.5735692977905273, "learning_rate": 9.831992573554454e-05, "loss": 0.0705, "step": 10290 }, { "grad_norm": 0.5504626035690308, "learning_rate": 9.831460883404128e-05, "loss": 0.053, "step": 10300 }, { "grad_norm": 0.6998856663703918, "learning_rate": 9.830928367689278e-05, "loss": 0.0598, "step": 10310 }, { "grad_norm": 0.45907577872276306, "learning_rate": 9.830395026500896e-05, "loss": 0.0659, "step": 10320 }, { "grad_norm": 0.6470319628715515, "learning_rate": 9.829860859930115e-05, "loss": 0.0614, "step": 10330 }, { "grad_norm": 0.5682451128959656, "learning_rate": 9.829325868068212e-05, "loss": 0.0706, "step": 10340 }, { "grad_norm": 0.558769941329956, "learning_rate": 9.8287900510066e-05, "loss": 0.0697, "step": 10350 }, { "grad_norm": 2.030223846435547, "learning_rate": 9.828253408836834e-05, "loss": 0.0648, "step": 10360 }, { "grad_norm": 0.5880056023597717, "learning_rate": 9.827715941650615e-05, "loss": 0.0611, "step": 10370 }, { "grad_norm": 0.4025496244430542, "learning_rate": 9.82717764953978e-05, "loss": 0.065, "step": 10380 }, { "grad_norm": 0.41581982374191284, "learning_rate": 9.826638532596308e-05, "loss": 0.0615, "step": 10390 }, { "grad_norm": 0.48700350522994995, "learning_rate": 9.82609859091232e-05, "loss": 0.061, "step": 10400 }, { "grad_norm": 0.4364685118198395, "learning_rate": 9.825557824580076e-05, "loss": 0.0632, "step": 10410 }, { "grad_norm": 0.5530269742012024, "learning_rate": 9.82501623369198e-05, "loss": 0.0697, "step": 10420 }, { "grad_norm": 0.4910651445388794, "learning_rate": 9.824473818340574e-05, "loss": 0.0529, "step": 10430 }, { "grad_norm": 0.5045240521430969, "learning_rate": 9.823930578618541e-05, "loss": 0.0559, "step": 10440 }, { "grad_norm": 0.5306631326675415, "learning_rate": 9.823386514618709e-05, "loss": 0.0619, "step": 10450 }, { "grad_norm": 0.6151505708694458, "learning_rate": 9.82284162643404e-05, "loss": 0.0623, "step": 10460 }, { "grad_norm": 0.46014001965522766, "learning_rate": 9.822295914157642e-05, "loss": 0.0607, "step": 10470 }, { "grad_norm": 0.4421336352825165, "learning_rate": 9.821749377882763e-05, "loss": 0.0596, "step": 10480 }, { "grad_norm": 0.40024083852767944, "learning_rate": 9.821202017702791e-05, "loss": 0.0524, "step": 10490 }, { "grad_norm": 0.4768838584423065, "learning_rate": 9.820653833711253e-05, "loss": 0.0538, "step": 10500 }, { "grad_norm": 0.5685042142868042, "learning_rate": 9.820104826001822e-05, "loss": 0.0648, "step": 10510 }, { "grad_norm": 0.4492112100124359, "learning_rate": 9.819554994668305e-05, "loss": 0.0563, "step": 10520 }, { "grad_norm": 0.5303248763084412, "learning_rate": 9.819004339804654e-05, "loss": 0.0592, "step": 10530 }, { "grad_norm": 0.47882869839668274, "learning_rate": 9.818452861504961e-05, "loss": 0.0524, "step": 10540 }, { "grad_norm": 0.4693322479724884, "learning_rate": 9.81790055986346e-05, "loss": 0.0527, "step": 10550 }, { "grad_norm": 0.38429006934165955, "learning_rate": 9.817347434974523e-05, "loss": 0.0534, "step": 10560 }, { "grad_norm": 0.448474258184433, "learning_rate": 9.816793486932664e-05, "loss": 0.0526, "step": 10570 }, { "grad_norm": 0.45591050386428833, "learning_rate": 9.816238715832538e-05, "loss": 0.0607, "step": 10580 }, { "grad_norm": 0.5490409731864929, "learning_rate": 9.815683121768939e-05, "loss": 0.0651, "step": 10590 }, { "grad_norm": 0.5514124631881714, "learning_rate": 9.815126704836804e-05, "loss": 0.0725, "step": 10600 }, { "grad_norm": 0.43900266289711, "learning_rate": 9.81456946513121e-05, "loss": 0.083, "step": 10610 }, { "grad_norm": 0.5464056730270386, "learning_rate": 9.814011402747373e-05, "loss": 0.0796, "step": 10620 }, { "grad_norm": 0.46694809198379517, "learning_rate": 9.813452517780651e-05, "loss": 0.0616, "step": 10630 }, { "grad_norm": 0.34054329991340637, "learning_rate": 9.81289281032654e-05, "loss": 0.0618, "step": 10640 }, { "grad_norm": 0.5180677771568298, "learning_rate": 9.812332280480683e-05, "loss": 0.0768, "step": 10650 }, { "grad_norm": 0.6234038472175598, "learning_rate": 9.811770928338854e-05, "loss": 0.0731, "step": 10660 }, { "grad_norm": 0.5410300493240356, "learning_rate": 9.811208753996979e-05, "loss": 0.0748, "step": 10670 }, { "grad_norm": 0.44313114881515503, "learning_rate": 9.810645757551113e-05, "loss": 0.0559, "step": 10680 }, { "grad_norm": 0.5034014582633972, "learning_rate": 9.810081939097459e-05, "loss": 0.0531, "step": 10690 }, { "grad_norm": 0.7994145750999451, "learning_rate": 9.809517298732356e-05, "loss": 0.0922, "step": 10700 }, { "grad_norm": 0.49790042638778687, "learning_rate": 9.80895183655229e-05, "loss": 0.0592, "step": 10710 }, { "grad_norm": 0.5861998200416565, "learning_rate": 9.808385552653877e-05, "loss": 0.0623, "step": 10720 }, { "grad_norm": 0.5271406173706055, "learning_rate": 9.807818447133886e-05, "loss": 0.0635, "step": 10730 }, { "grad_norm": 0.7715872526168823, "learning_rate": 9.807250520089215e-05, "loss": 0.0664, "step": 10740 }, { "grad_norm": 0.6546924114227295, "learning_rate": 9.806681771616908e-05, "loss": 0.0707, "step": 10750 }, { "grad_norm": 0.4986235201358795, "learning_rate": 9.80611220181415e-05, "loss": 0.0715, "step": 10760 }, { "grad_norm": 0.5711162090301514, "learning_rate": 9.805541810778264e-05, "loss": 0.0648, "step": 10770 }, { "grad_norm": 0.6509902477264404, "learning_rate": 9.804970598606716e-05, "loss": 0.0746, "step": 10780 }, { "grad_norm": 0.5702444314956665, "learning_rate": 9.804398565397106e-05, "loss": 0.0697, "step": 10790 }, { "grad_norm": 0.3671962320804596, "learning_rate": 9.803825711247183e-05, "loss": 0.0695, "step": 10800 }, { "grad_norm": 0.4794487953186035, "learning_rate": 9.803252036254831e-05, "loss": 0.0631, "step": 10810 }, { "grad_norm": 0.8290128111839294, "learning_rate": 9.802677540518076e-05, "loss": 0.073, "step": 10820 }, { "grad_norm": 0.5064197778701782, "learning_rate": 9.802102224135081e-05, "loss": 0.0678, "step": 10830 }, { "grad_norm": 0.5619979500770569, "learning_rate": 9.801526087204155e-05, "loss": 0.0843, "step": 10840 }, { "grad_norm": 0.5372827649116516, "learning_rate": 9.800949129823743e-05, "loss": 0.0675, "step": 10850 }, { "grad_norm": 0.4684506058692932, "learning_rate": 9.80037135209243e-05, "loss": 0.0559, "step": 10860 }, { "grad_norm": 0.5911660194396973, "learning_rate": 9.799792754108946e-05, "loss": 0.0754, "step": 10870 }, { "grad_norm": 0.3971153795719147, "learning_rate": 9.799213335972152e-05, "loss": 0.0654, "step": 10880 }, { "grad_norm": 0.40710219740867615, "learning_rate": 9.798633097781058e-05, "loss": 0.0695, "step": 10890 }, { "grad_norm": 0.3865611255168915, "learning_rate": 9.79805203963481e-05, "loss": 0.0659, "step": 10900 }, { "grad_norm": 0.5934469103813171, "learning_rate": 9.797470161632697e-05, "loss": 0.0623, "step": 10910 }, { "grad_norm": 0.5271896123886108, "learning_rate": 9.796887463874145e-05, "loss": 0.0693, "step": 10920 }, { "grad_norm": 0.342693567276001, "learning_rate": 9.796303946458718e-05, "loss": 0.0656, "step": 10930 }, { "grad_norm": 0.5189915299415588, "learning_rate": 9.795719609486127e-05, "loss": 0.0734, "step": 10940 }, { "grad_norm": 0.6003278493881226, "learning_rate": 9.795134453056219e-05, "loss": 0.0511, "step": 10950 }, { "grad_norm": 0.43814617395401, "learning_rate": 9.794548477268979e-05, "loss": 0.0516, "step": 10960 }, { "grad_norm": 0.5854668021202087, "learning_rate": 9.793961682224537e-05, "loss": 0.05, "step": 10970 }, { "grad_norm": 0.4467281401157379, "learning_rate": 9.793374068023156e-05, "loss": 0.0589, "step": 10980 }, { "grad_norm": 0.5659717321395874, "learning_rate": 9.792785634765247e-05, "loss": 0.0621, "step": 10990 }, { "grad_norm": 0.4417332112789154, "learning_rate": 9.792196382551357e-05, "loss": 0.0586, "step": 11000 }, { "grad_norm": 0.40620356798171997, "learning_rate": 9.791606311482171e-05, "loss": 0.0583, "step": 11010 }, { "grad_norm": 0.6857539415359497, "learning_rate": 9.791015421658518e-05, "loss": 0.0604, "step": 11020 }, { "grad_norm": 0.5216180086135864, "learning_rate": 9.790423713181362e-05, "loss": 0.0604, "step": 11030 }, { "grad_norm": 0.4614591598510742, "learning_rate": 9.789831186151814e-05, "loss": 0.0638, "step": 11040 }, { "grad_norm": 0.4823971688747406, "learning_rate": 9.789237840671118e-05, "loss": 0.0695, "step": 11050 }, { "grad_norm": 0.48072659969329834, "learning_rate": 9.78864367684066e-05, "loss": 0.063, "step": 11060 }, { "grad_norm": 0.5222879648208618, "learning_rate": 9.788048694761968e-05, "loss": 0.0567, "step": 11070 }, { "grad_norm": 0.48093342781066895, "learning_rate": 9.787452894536709e-05, "loss": 0.0574, "step": 11080 }, { "grad_norm": 0.36419880390167236, "learning_rate": 9.786856276266685e-05, "loss": 0.0565, "step": 11090 }, { "grad_norm": 0.49150896072387695, "learning_rate": 9.786258840053845e-05, "loss": 0.0659, "step": 11100 }, { "grad_norm": 0.49194374680519104, "learning_rate": 9.785660586000273e-05, "loss": 0.0652, "step": 11110 }, { "grad_norm": 0.5811507105827332, "learning_rate": 9.785061514208196e-05, "loss": 0.0661, "step": 11120 }, { "grad_norm": 0.5248380899429321, "learning_rate": 9.784461624779977e-05, "loss": 0.071, "step": 11130 }, { "grad_norm": 0.7410138249397278, "learning_rate": 9.783860917818123e-05, "loss": 0.0595, "step": 11140 }, { "grad_norm": 0.4673278033733368, "learning_rate": 9.783259393425277e-05, "loss": 0.0453, "step": 11150 }, { "grad_norm": 0.4810558557510376, "learning_rate": 9.782657051704221e-05, "loss": 0.0482, "step": 11160 }, { "grad_norm": 0.667806088924408, "learning_rate": 9.782053892757883e-05, "loss": 0.056, "step": 11170 }, { "grad_norm": 0.5594167113304138, "learning_rate": 9.781449916689324e-05, "loss": 0.0615, "step": 11180 }, { "grad_norm": 0.4845450222492218, "learning_rate": 9.780845123601746e-05, "loss": 0.058, "step": 11190 }, { "grad_norm": 0.4271129369735718, "learning_rate": 9.780239513598492e-05, "loss": 0.0641, "step": 11200 }, { "grad_norm": 0.5777284502983093, "learning_rate": 9.779633086783047e-05, "loss": 0.0647, "step": 11210 }, { "grad_norm": 0.46274641156196594, "learning_rate": 9.779025843259031e-05, "loss": 0.0575, "step": 11220 }, { "grad_norm": 0.4936058521270752, "learning_rate": 9.778417783130204e-05, "loss": 0.0694, "step": 11230 }, { "grad_norm": 0.34586435556411743, "learning_rate": 9.777808906500468e-05, "loss": 0.0573, "step": 11240 }, { "grad_norm": 0.6394819021224976, "learning_rate": 9.777199213473862e-05, "loss": 0.053, "step": 11250 }, { "grad_norm": 0.4252638816833496, "learning_rate": 9.77658870415457e-05, "loss": 0.0508, "step": 11260 }, { "grad_norm": 0.4537487328052521, "learning_rate": 9.775977378646906e-05, "loss": 0.0645, "step": 11270 }, { "grad_norm": 0.8303909301757812, "learning_rate": 9.775365237055331e-05, "loss": 0.074, "step": 11280 }, { "grad_norm": 0.5070237517356873, "learning_rate": 9.774752279484445e-05, "loss": 0.0573, "step": 11290 }, { "grad_norm": 0.47700175642967224, "learning_rate": 9.774138506038984e-05, "loss": 0.0547, "step": 11300 }, { "grad_norm": 0.3979994058609009, "learning_rate": 9.773523916823826e-05, "loss": 0.0589, "step": 11310 }, { "grad_norm": 0.37959024310112, "learning_rate": 9.772908511943986e-05, "loss": 0.051, "step": 11320 }, { "grad_norm": 0.4317716956138611, "learning_rate": 9.77229229150462e-05, "loss": 0.051, "step": 11330 }, { "grad_norm": 0.5356541872024536, "learning_rate": 9.771675255611024e-05, "loss": 0.0433, "step": 11340 }, { "grad_norm": 0.45107510685920715, "learning_rate": 9.771057404368632e-05, "loss": 0.0653, "step": 11350 }, { "grad_norm": 0.4607129395008087, "learning_rate": 9.770438737883018e-05, "loss": 0.0502, "step": 11360 }, { "grad_norm": 0.5569575428962708, "learning_rate": 9.769819256259898e-05, "loss": 0.045, "step": 11370 }, { "grad_norm": 0.3577004671096802, "learning_rate": 9.769198959605119e-05, "loss": 0.0502, "step": 11380 }, { "grad_norm": 0.43620574474334717, "learning_rate": 9.768577848024678e-05, "loss": 0.0637, "step": 11390 }, { "grad_norm": 0.48993298411369324, "learning_rate": 9.767955921624702e-05, "loss": 0.0674, "step": 11400 }, { "grad_norm": 0.5554031729698181, "learning_rate": 9.767333180511465e-05, "loss": 0.055, "step": 11410 }, { "grad_norm": 0.4432331323623657, "learning_rate": 9.766709624791373e-05, "loss": 0.0692, "step": 11420 }, { "grad_norm": 0.4994751214981079, "learning_rate": 9.766085254570975e-05, "loss": 0.057, "step": 11430 }, { "grad_norm": 0.5512555241584778, "learning_rate": 9.76546006995696e-05, "loss": 0.0553, "step": 11440 }, { "grad_norm": 0.5333337783813477, "learning_rate": 9.764834071056155e-05, "loss": 0.0554, "step": 11450 }, { "grad_norm": 0.4658060073852539, "learning_rate": 9.764207257975526e-05, "loss": 0.0564, "step": 11460 }, { "grad_norm": 0.7517643570899963, "learning_rate": 9.763579630822179e-05, "loss": 0.0549, "step": 11470 }, { "grad_norm": 0.4828474521636963, "learning_rate": 9.762951189703356e-05, "loss": 0.0502, "step": 11480 }, { "grad_norm": 0.6132215261459351, "learning_rate": 9.762321934726442e-05, "loss": 0.0606, "step": 11490 }, { "grad_norm": 0.46485763788223267, "learning_rate": 9.761691865998959e-05, "loss": 0.06, "step": 11500 }, { "grad_norm": 0.5444358587265015, "learning_rate": 9.76106098362857e-05, "loss": 0.0517, "step": 11510 }, { "grad_norm": 0.4517616629600525, "learning_rate": 9.760429287723072e-05, "loss": 0.063, "step": 11520 }, { "grad_norm": 0.4137392044067383, "learning_rate": 9.759796778390406e-05, "loss": 0.0537, "step": 11530 }, { "grad_norm": 0.4398830235004425, "learning_rate": 9.759163455738653e-05, "loss": 0.057, "step": 11540 }, { "grad_norm": 0.5373238325119019, "learning_rate": 9.75852931987603e-05, "loss": 0.056, "step": 11550 }, { "grad_norm": 0.3153674602508545, "learning_rate": 9.757894370910891e-05, "loss": 0.0539, "step": 11560 }, { "grad_norm": 0.6285452842712402, "learning_rate": 9.757258608951733e-05, "loss": 0.0516, "step": 11570 }, { "grad_norm": 0.48742905259132385, "learning_rate": 9.75662203410719e-05, "loss": 0.0599, "step": 11580 }, { "grad_norm": 0.5280252695083618, "learning_rate": 9.755984646486034e-05, "loss": 0.0626, "step": 11590 }, { "grad_norm": 0.5181785225868225, "learning_rate": 9.75534644619718e-05, "loss": 0.0542, "step": 11600 }, { "grad_norm": 0.42221519351005554, "learning_rate": 9.754707433349676e-05, "loss": 0.0619, "step": 11610 }, { "grad_norm": 0.6373863220214844, "learning_rate": 9.754067608052715e-05, "loss": 0.0674, "step": 11620 }, { "grad_norm": 0.520429253578186, "learning_rate": 9.753426970415622e-05, "loss": 0.0825, "step": 11630 }, { "grad_norm": 0.37214142084121704, "learning_rate": 9.752785520547868e-05, "loss": 0.0542, "step": 11640 }, { "grad_norm": 0.5693328976631165, "learning_rate": 9.752143258559056e-05, "loss": 0.0587, "step": 11650 }, { "grad_norm": 0.529677152633667, "learning_rate": 9.751500184558933e-05, "loss": 0.0585, "step": 11660 }, { "grad_norm": 0.3418262004852295, "learning_rate": 9.750856298657383e-05, "loss": 0.0523, "step": 11670 }, { "grad_norm": 0.5116968154907227, "learning_rate": 9.750211600964428e-05, "loss": 0.0653, "step": 11680 }, { "grad_norm": 0.5574443340301514, "learning_rate": 9.749566091590226e-05, "loss": 0.0503, "step": 11690 }, { "grad_norm": 0.4272407293319702, "learning_rate": 9.748919770645083e-05, "loss": 0.0597, "step": 11700 }, { "grad_norm": 0.4830801486968994, "learning_rate": 9.748272638239432e-05, "loss": 0.0595, "step": 11710 }, { "grad_norm": 0.5878814458847046, "learning_rate": 9.747624694483855e-05, "loss": 0.0559, "step": 11720 }, { "grad_norm": 0.4466489255428314, "learning_rate": 9.746975939489065e-05, "loss": 0.0463, "step": 11730 }, { "grad_norm": 0.5342978835105896, "learning_rate": 9.746326373365918e-05, "loss": 0.0573, "step": 11740 }, { "grad_norm": 0.48733678460121155, "learning_rate": 9.745675996225403e-05, "loss": 0.0551, "step": 11750 }, { "grad_norm": 0.5692518353462219, "learning_rate": 9.745024808178657e-05, "loss": 0.0498, "step": 11760 }, { "grad_norm": 0.35609525442123413, "learning_rate": 9.744372809336947e-05, "loss": 0.0602, "step": 11770 }, { "grad_norm": 0.43709608912467957, "learning_rate": 9.743719999811682e-05, "loss": 0.0539, "step": 11780 }, { "grad_norm": 0.5777975916862488, "learning_rate": 9.743066379714412e-05, "loss": 0.0498, "step": 11790 }, { "grad_norm": 0.3627598285675049, "learning_rate": 9.74241194915682e-05, "loss": 0.0488, "step": 11800 }, { "grad_norm": 0.5197979807853699, "learning_rate": 9.741756708250731e-05, "loss": 0.0668, "step": 11810 }, { "grad_norm": 0.3818711042404175, "learning_rate": 9.741100657108109e-05, "loss": 0.0579, "step": 11820 }, { "grad_norm": 0.49676525592803955, "learning_rate": 9.740443795841054e-05, "loss": 0.057, "step": 11830 }, { "grad_norm": 0.3862050175666809, "learning_rate": 9.739786124561805e-05, "loss": 0.0462, "step": 11840 }, { "grad_norm": 0.3982596695423126, "learning_rate": 9.73912764338274e-05, "loss": 0.0458, "step": 11850 }, { "grad_norm": 0.37072253227233887, "learning_rate": 9.738468352416377e-05, "loss": 0.0603, "step": 11860 }, { "grad_norm": 0.5161729454994202, "learning_rate": 9.737808251775369e-05, "loss": 0.0514, "step": 11870 }, { "grad_norm": 0.3680870532989502, "learning_rate": 9.737147341572512e-05, "loss": 0.0709, "step": 11880 }, { "grad_norm": 0.39176318049430847, "learning_rate": 9.736485621920735e-05, "loss": 0.053, "step": 11890 }, { "grad_norm": 0.47503355145454407, "learning_rate": 9.735823092933108e-05, "loss": 0.0579, "step": 11900 }, { "grad_norm": 0.39409154653549194, "learning_rate": 9.735159754722838e-05, "loss": 0.0553, "step": 11910 }, { "grad_norm": 0.41000762581825256, "learning_rate": 9.734495607403275e-05, "loss": 0.055, "step": 11920 }, { "grad_norm": 0.4548577070236206, "learning_rate": 9.733830651087901e-05, "loss": 0.0522, "step": 11930 }, { "grad_norm": 0.4612186551094055, "learning_rate": 9.733164885890338e-05, "loss": 0.0536, "step": 11940 }, { "grad_norm": 0.34663787484169006, "learning_rate": 9.732498311924349e-05, "loss": 0.0459, "step": 11950 }, { "grad_norm": 0.5703560709953308, "learning_rate": 9.731830929303833e-05, "loss": 0.0551, "step": 11960 }, { "grad_norm": 0.6950038075447083, "learning_rate": 9.731162738142827e-05, "loss": 0.0542, "step": 11970 }, { "grad_norm": 0.5063934326171875, "learning_rate": 9.730493738555506e-05, "loss": 0.0546, "step": 11980 }, { "grad_norm": 0.4056486487388611, "learning_rate": 9.729823930656186e-05, "loss": 0.0513, "step": 11990 }, { "grad_norm": 0.5020638704299927, "learning_rate": 9.729153314559316e-05, "loss": 0.0724, "step": 12000 }, { "grad_norm": 0.5172890424728394, "learning_rate": 9.728481890379486e-05, "loss": 0.0597, "step": 12010 }, { "grad_norm": 0.4624101221561432, "learning_rate": 9.727809658231428e-05, "loss": 0.05, "step": 12020 }, { "grad_norm": 0.6059184670448303, "learning_rate": 9.727136618230003e-05, "loss": 0.0542, "step": 12030 }, { "grad_norm": 0.5304904580116272, "learning_rate": 9.726462770490219e-05, "loss": 0.06, "step": 12040 }, { "grad_norm": 0.506965160369873, "learning_rate": 9.725788115127214e-05, "loss": 0.0604, "step": 12050 }, { "grad_norm": 0.3702981173992157, "learning_rate": 9.725112652256274e-05, "loss": 0.0556, "step": 12060 }, { "grad_norm": 0.5230512619018555, "learning_rate": 9.724436381992812e-05, "loss": 0.082, "step": 12070 }, { "grad_norm": 0.5980575084686279, "learning_rate": 9.723759304452387e-05, "loss": 0.0599, "step": 12080 }, { "grad_norm": 0.47430941462516785, "learning_rate": 9.72308141975069e-05, "loss": 0.0448, "step": 12090 }, { "grad_norm": 0.5041361451148987, "learning_rate": 9.722402728003557e-05, "loss": 0.0594, "step": 12100 }, { "grad_norm": 0.6276789307594299, "learning_rate": 9.721723229326953e-05, "loss": 0.0549, "step": 12110 }, { "grad_norm": 0.5259868502616882, "learning_rate": 9.721042923836992e-05, "loss": 0.047, "step": 12120 }, { "grad_norm": 0.4537011682987213, "learning_rate": 9.720361811649914e-05, "loss": 0.055, "step": 12130 }, { "grad_norm": 0.3610376715660095, "learning_rate": 9.719679892882106e-05, "loss": 0.0448, "step": 12140 }, { "grad_norm": 0.44537970423698425, "learning_rate": 9.718997167650085e-05, "loss": 0.0474, "step": 12150 }, { "grad_norm": 0.42291030287742615, "learning_rate": 9.718313636070515e-05, "loss": 0.046, "step": 12160 }, { "grad_norm": 0.5402450561523438, "learning_rate": 9.717629298260192e-05, "loss": 0.0447, "step": 12170 }, { "grad_norm": 0.5687583088874817, "learning_rate": 9.716944154336047e-05, "loss": 0.0623, "step": 12180 }, { "grad_norm": 0.48539456725120544, "learning_rate": 9.716258204415157e-05, "loss": 0.0537, "step": 12190 }, { "grad_norm": 0.478407621383667, "learning_rate": 9.715571448614728e-05, "loss": 0.0641, "step": 12200 }, { "grad_norm": 0.42163434624671936, "learning_rate": 9.71488388705211e-05, "loss": 0.075, "step": 12210 }, { "grad_norm": 0.5546137094497681, "learning_rate": 9.714195519844788e-05, "loss": 0.0573, "step": 12220 }, { "grad_norm": 0.4331255257129669, "learning_rate": 9.713506347110386e-05, "loss": 0.0574, "step": 12230 }, { "grad_norm": 0.5484188795089722, "learning_rate": 9.712816368966662e-05, "loss": 0.0556, "step": 12240 }, { "grad_norm": 0.4447690546512604, "learning_rate": 9.712125585531517e-05, "loss": 0.07, "step": 12250 }, { "grad_norm": 0.6103976964950562, "learning_rate": 9.711433996922988e-05, "loss": 0.0622, "step": 12260 }, { "grad_norm": 0.3849402964115143, "learning_rate": 9.710741603259245e-05, "loss": 0.0502, "step": 12270 }, { "grad_norm": 0.34508344531059265, "learning_rate": 9.710048404658603e-05, "loss": 0.0525, "step": 12280 }, { "grad_norm": 0.4568350613117218, "learning_rate": 9.709354401239508e-05, "loss": 0.0645, "step": 12290 }, { "grad_norm": 0.4897104501724243, "learning_rate": 9.708659593120546e-05, "loss": 0.0515, "step": 12300 }, { "grad_norm": 0.4445017874240875, "learning_rate": 9.707963980420443e-05, "loss": 0.0583, "step": 12310 }, { "grad_norm": 0.7003058195114136, "learning_rate": 9.707267563258058e-05, "loss": 0.063, "step": 12320 }, { "grad_norm": 0.5563026070594788, "learning_rate": 9.70657034175239e-05, "loss": 0.0583, "step": 12330 }, { "grad_norm": 0.3964361846446991, "learning_rate": 9.705872316022577e-05, "loss": 0.0562, "step": 12340 }, { "grad_norm": 0.4545851945877075, "learning_rate": 9.705173486187891e-05, "loss": 0.0552, "step": 12350 }, { "grad_norm": 0.5725316405296326, "learning_rate": 9.704473852367741e-05, "loss": 0.0618, "step": 12360 }, { "grad_norm": 0.40881022810935974, "learning_rate": 9.70377341468168e-05, "loss": 0.0547, "step": 12370 }, { "grad_norm": 0.5663781762123108, "learning_rate": 9.703072173249389e-05, "loss": 0.0576, "step": 12380 }, { "grad_norm": 0.5837775468826294, "learning_rate": 9.702370128190693e-05, "loss": 0.0734, "step": 12390 }, { "grad_norm": 0.5877337455749512, "learning_rate": 9.701667279625552e-05, "loss": 0.0616, "step": 12400 }, { "grad_norm": 0.5482819080352783, "learning_rate": 9.700963627674065e-05, "loss": 0.0649, "step": 12410 }, { "grad_norm": 0.5286867022514343, "learning_rate": 9.700259172456466e-05, "loss": 0.0514, "step": 12420 }, { "grad_norm": 0.5186893939971924, "learning_rate": 9.699553914093124e-05, "loss": 0.067, "step": 12430 }, { "grad_norm": 0.432779461145401, "learning_rate": 9.698847852704553e-05, "loss": 0.0617, "step": 12440 }, { "grad_norm": 0.4586676359176636, "learning_rate": 9.6981409884114e-05, "loss": 0.0457, "step": 12450 }, { "grad_norm": 0.4769275486469269, "learning_rate": 9.697433321334443e-05, "loss": 0.0584, "step": 12460 }, { "grad_norm": 0.38923248648643494, "learning_rate": 9.696724851594607e-05, "loss": 0.0478, "step": 12470 }, { "grad_norm": 0.3574082553386688, "learning_rate": 9.696015579312952e-05, "loss": 0.0466, "step": 12480 }, { "grad_norm": 0.3520236909389496, "learning_rate": 9.695305504610668e-05, "loss": 0.0464, "step": 12490 }, { "grad_norm": 0.3642467260360718, "learning_rate": 9.694594627609092e-05, "loss": 0.0576, "step": 12500 }, { "grad_norm": 0.4793960154056549, "learning_rate": 9.693882948429691e-05, "loss": 0.0623, "step": 12510 }, { "grad_norm": 0.516390323638916, "learning_rate": 9.693170467194071e-05, "loss": 0.0503, "step": 12520 }, { "grad_norm": 0.48768457770347595, "learning_rate": 9.692457184023977e-05, "loss": 0.0483, "step": 12530 }, { "grad_norm": 0.4096086621284485, "learning_rate": 9.691743099041291e-05, "loss": 0.0584, "step": 12540 }, { "grad_norm": 0.6597204804420471, "learning_rate": 9.691028212368027e-05, "loss": 0.0538, "step": 12550 }, { "grad_norm": 0.465901255607605, "learning_rate": 9.690312524126342e-05, "loss": 0.0548, "step": 12560 }, { "grad_norm": 0.4719499349594116, "learning_rate": 9.689596034438527e-05, "loss": 0.0585, "step": 12570 }, { "grad_norm": 0.4826265275478363, "learning_rate": 9.688878743427012e-05, "loss": 0.0536, "step": 12580 }, { "grad_norm": 0.557027280330658, "learning_rate": 9.688160651214359e-05, "loss": 0.0652, "step": 12590 }, { "grad_norm": 0.4347469210624695, "learning_rate": 9.687441757923273e-05, "loss": 0.0492, "step": 12600 }, { "grad_norm": 0.5799979567527771, "learning_rate": 9.68672206367659e-05, "loss": 0.059, "step": 12610 }, { "grad_norm": 0.5237774848937988, "learning_rate": 9.686001568597291e-05, "loss": 0.0492, "step": 12620 }, { "grad_norm": 0.38828155398368835, "learning_rate": 9.685280272808486e-05, "loss": 0.0468, "step": 12630 }, { "grad_norm": 0.4682970941066742, "learning_rate": 9.684558176433424e-05, "loss": 0.0659, "step": 12640 }, { "grad_norm": 0.48954302072525024, "learning_rate": 9.683835279595495e-05, "loss": 0.0539, "step": 12650 }, { "grad_norm": 0.630893886089325, "learning_rate": 9.683111582418216e-05, "loss": 0.0697, "step": 12660 }, { "grad_norm": 0.44289129972457886, "learning_rate": 9.682387085025254e-05, "loss": 0.0596, "step": 12670 }, { "grad_norm": 0.5100993514060974, "learning_rate": 9.681661787540401e-05, "loss": 0.0624, "step": 12680 }, { "grad_norm": 0.45065996050834656, "learning_rate": 9.680935690087593e-05, "loss": 0.0631, "step": 12690 }, { "grad_norm": 0.5045967102050781, "learning_rate": 9.680208792790901e-05, "loss": 0.0579, "step": 12700 }, { "grad_norm": 0.5223396420478821, "learning_rate": 9.679481095774529e-05, "loss": 0.0545, "step": 12710 }, { "grad_norm": 0.5233434438705444, "learning_rate": 9.678752599162822e-05, "loss": 0.0577, "step": 12720 }, { "grad_norm": 0.5678813457489014, "learning_rate": 9.678023303080259e-05, "loss": 0.0486, "step": 12730 }, { "grad_norm": 0.3995141386985779, "learning_rate": 9.677293207651459e-05, "loss": 0.0453, "step": 12740 }, { "grad_norm": 0.3754066228866577, "learning_rate": 9.676562313001173e-05, "loss": 0.0478, "step": 12750 }, { "grad_norm": 0.6180054545402527, "learning_rate": 9.675830619254293e-05, "loss": 0.061, "step": 12760 }, { "grad_norm": 0.4801599681377411, "learning_rate": 9.675098126535843e-05, "loss": 0.0615, "step": 12770 }, { "grad_norm": 0.6788285970687866, "learning_rate": 9.674364834970988e-05, "loss": 0.064, "step": 12780 }, { "grad_norm": 0.4510735869407654, "learning_rate": 9.673630744685028e-05, "loss": 0.0482, "step": 12790 }, { "grad_norm": 0.35843586921691895, "learning_rate": 9.672895855803397e-05, "loss": 0.0545, "step": 12800 }, { "grad_norm": 0.4685962498188019, "learning_rate": 9.672160168451667e-05, "loss": 0.0496, "step": 12810 }, { "grad_norm": 0.40749338269233704, "learning_rate": 9.671423682755549e-05, "loss": 0.0594, "step": 12820 }, { "grad_norm": 0.6025264859199524, "learning_rate": 9.670686398840888e-05, "loss": 0.0636, "step": 12830 }, { "grad_norm": 0.3916870653629303, "learning_rate": 9.669948316833664e-05, "loss": 0.0522, "step": 12840 }, { "grad_norm": 0.4433448016643524, "learning_rate": 9.669209436859997e-05, "loss": 0.0562, "step": 12850 }, { "grad_norm": 0.4361985921859741, "learning_rate": 9.66846975904614e-05, "loss": 0.0656, "step": 12860 }, { "grad_norm": 0.5816381573677063, "learning_rate": 9.667729283518483e-05, "loss": 0.0554, "step": 12870 }, { "grad_norm": 0.5315982103347778, "learning_rate": 9.666988010403557e-05, "loss": 0.0415, "step": 12880 }, { "grad_norm": 0.4047679007053375, "learning_rate": 9.66624593982802e-05, "loss": 0.047, "step": 12890 }, { "grad_norm": 0.43748489022254944, "learning_rate": 9.665503071918675e-05, "loss": 0.0586, "step": 12900 }, { "grad_norm": 0.43019774556159973, "learning_rate": 9.664759406802456e-05, "loss": 0.05, "step": 12910 }, { "grad_norm": 0.414287805557251, "learning_rate": 9.664014944606437e-05, "loss": 0.0453, "step": 12920 }, { "grad_norm": 0.42771822214126587, "learning_rate": 9.663269685457822e-05, "loss": 0.0542, "step": 12930 }, { "grad_norm": 0.45109549164772034, "learning_rate": 9.662523629483962e-05, "loss": 0.0707, "step": 12940 }, { "grad_norm": 0.400563508272171, "learning_rate": 9.661776776812333e-05, "loss": 0.0447, "step": 12950 }, { "grad_norm": 0.5607568025588989, "learning_rate": 9.661029127570553e-05, "loss": 0.0511, "step": 12960 }, { "grad_norm": 0.5530173778533936, "learning_rate": 9.660280681886373e-05, "loss": 0.0566, "step": 12970 }, { "grad_norm": 0.6149089932441711, "learning_rate": 9.659531439887685e-05, "loss": 0.0482, "step": 12980 }, { "grad_norm": 0.4431595504283905, "learning_rate": 9.658781401702511e-05, "loss": 0.0507, "step": 12990 }, { "grad_norm": 0.510554850101471, "learning_rate": 9.658030567459015e-05, "loss": 0.055, "step": 13000 }, { "grad_norm": 0.5935876965522766, "learning_rate": 9.65727893728549e-05, "loss": 0.0676, "step": 13010 }, { "grad_norm": 0.6076684594154358, "learning_rate": 9.656526511310375e-05, "loss": 0.0629, "step": 13020 }, { "grad_norm": 0.48992785811424255, "learning_rate": 9.655773289662233e-05, "loss": 0.0675, "step": 13030 }, { "grad_norm": 0.47305306792259216, "learning_rate": 9.655019272469772e-05, "loss": 0.0576, "step": 13040 }, { "grad_norm": 0.5046154260635376, "learning_rate": 9.654264459861832e-05, "loss": 0.0422, "step": 13050 }, { "grad_norm": 0.4922158122062683, "learning_rate": 9.653508851967391e-05, "loss": 0.0576, "step": 13060 }, { "grad_norm": 0.5520263314247131, "learning_rate": 9.65275244891556e-05, "loss": 0.0682, "step": 13070 }, { "grad_norm": 0.41680026054382324, "learning_rate": 9.651995250835591e-05, "loss": 0.0516, "step": 13080 }, { "grad_norm": 0.3990180492401123, "learning_rate": 9.651237257856862e-05, "loss": 0.0554, "step": 13090 }, { "grad_norm": 0.5023607015609741, "learning_rate": 9.6504784701089e-05, "loss": 0.0549, "step": 13100 }, { "grad_norm": 0.4439077079296112, "learning_rate": 9.649718887721357e-05, "loss": 0.0512, "step": 13110 }, { "grad_norm": 0.40493202209472656, "learning_rate": 9.648958510824028e-05, "loss": 0.0544, "step": 13120 }, { "grad_norm": 0.4269993305206299, "learning_rate": 9.648197339546837e-05, "loss": 0.0601, "step": 13130 }, { "grad_norm": 0.4799160063266754, "learning_rate": 9.647435374019851e-05, "loss": 0.0603, "step": 13140 }, { "grad_norm": 0.623820960521698, "learning_rate": 9.646672614373266e-05, "loss": 0.0595, "step": 13150 }, { "grad_norm": 0.5536771416664124, "learning_rate": 9.645909060737418e-05, "loss": 0.0507, "step": 13160 }, { "grad_norm": 0.3986974358558655, "learning_rate": 9.645144713242778e-05, "loss": 0.0571, "step": 13170 }, { "grad_norm": 0.5171501636505127, "learning_rate": 9.64437957201995e-05, "loss": 0.0536, "step": 13180 }, { "grad_norm": 0.4365232288837433, "learning_rate": 9.643613637199678e-05, "loss": 0.0582, "step": 13190 }, { "grad_norm": 0.5495362877845764, "learning_rate": 9.642846908912839e-05, "loss": 0.0623, "step": 13200 }, { "grad_norm": 0.4990967810153961, "learning_rate": 9.642079387290444e-05, "loss": 0.0486, "step": 13210 }, { "grad_norm": 0.4153241515159607, "learning_rate": 9.641311072463644e-05, "loss": 0.0554, "step": 13220 }, { "grad_norm": 0.5786349773406982, "learning_rate": 9.640541964563722e-05, "loss": 0.0549, "step": 13230 }, { "grad_norm": 0.5444525480270386, "learning_rate": 9.639772063722096e-05, "loss": 0.0599, "step": 13240 }, { "grad_norm": 0.40459388494491577, "learning_rate": 9.639001370070324e-05, "loss": 0.0572, "step": 13250 }, { "grad_norm": 0.569763720035553, "learning_rate": 9.638229883740095e-05, "loss": 0.0604, "step": 13260 }, { "grad_norm": 0.5143391489982605, "learning_rate": 9.637457604863233e-05, "loss": 0.0539, "step": 13270 }, { "grad_norm": 0.46817687153816223, "learning_rate": 9.636684533571703e-05, "loss": 0.0499, "step": 13280 }, { "grad_norm": 0.439614862203598, "learning_rate": 9.635910669997599e-05, "loss": 0.0432, "step": 13290 }, { "grad_norm": 0.5827438831329346, "learning_rate": 9.635136014273154e-05, "loss": 0.0459, "step": 13300 }, { "grad_norm": 0.45343247056007385, "learning_rate": 9.634360566530735e-05, "loss": 0.0592, "step": 13310 }, { "grad_norm": 0.5507816672325134, "learning_rate": 9.633584326902845e-05, "loss": 0.0614, "step": 13320 }, { "grad_norm": 0.5542444586753845, "learning_rate": 9.632807295522124e-05, "loss": 0.061, "step": 13330 }, { "grad_norm": 0.5439581871032715, "learning_rate": 9.632029472521342e-05, "loss": 0.0522, "step": 13340 }, { "grad_norm": 0.5191962718963623, "learning_rate": 9.631250858033409e-05, "loss": 0.0532, "step": 13350 }, { "grad_norm": 0.4562129080295563, "learning_rate": 9.630471452191371e-05, "loss": 0.0474, "step": 13360 }, { "grad_norm": 0.40184664726257324, "learning_rate": 9.629691255128405e-05, "loss": 0.0632, "step": 13370 }, { "grad_norm": 0.6320075988769531, "learning_rate": 9.628910266977825e-05, "loss": 0.0493, "step": 13380 }, { "grad_norm": 0.45649853348731995, "learning_rate": 9.628128487873083e-05, "loss": 0.0652, "step": 13390 }, { "grad_norm": 0.4819071292877197, "learning_rate": 9.627345917947761e-05, "loss": 0.0669, "step": 13400 }, { "grad_norm": 0.4172893464565277, "learning_rate": 9.626562557335579e-05, "loss": 0.0593, "step": 13410 }, { "grad_norm": 0.5385345816612244, "learning_rate": 9.625778406170393e-05, "loss": 0.0524, "step": 13420 }, { "grad_norm": 0.5747860074043274, "learning_rate": 9.624993464586193e-05, "loss": 0.0541, "step": 13430 }, { "grad_norm": 0.5491486191749573, "learning_rate": 9.624207732717105e-05, "loss": 0.0505, "step": 13440 }, { "grad_norm": 0.616737961769104, "learning_rate": 9.623421210697386e-05, "loss": 0.0524, "step": 13450 }, { "grad_norm": 0.4431520998477936, "learning_rate": 9.622633898661434e-05, "loss": 0.0575, "step": 13460 }, { "grad_norm": 0.5652044415473938, "learning_rate": 9.621845796743778e-05, "loss": 0.0654, "step": 13470 }, { "grad_norm": 0.32577648758888245, "learning_rate": 9.621056905079082e-05, "loss": 0.0659, "step": 13480 }, { "grad_norm": 0.5193151831626892, "learning_rate": 9.620267223802149e-05, "loss": 0.067, "step": 13490 }, { "grad_norm": 0.41401195526123047, "learning_rate": 9.619476753047911e-05, "loss": 0.0483, "step": 13500 }, { "grad_norm": 0.45515888929367065, "learning_rate": 9.618685492951438e-05, "loss": 0.0663, "step": 13510 }, { "grad_norm": 0.357828825712204, "learning_rate": 9.617893443647938e-05, "loss": 0.0607, "step": 13520 }, { "grad_norm": 0.657977283000946, "learning_rate": 9.617100605272746e-05, "loss": 0.0648, "step": 13530 }, { "grad_norm": 0.39726343750953674, "learning_rate": 9.616306977961338e-05, "loss": 0.0591, "step": 13540 }, { "grad_norm": 0.4627246558666229, "learning_rate": 9.615512561849326e-05, "loss": 0.0459, "step": 13550 }, { "grad_norm": 0.4715941846370697, "learning_rate": 9.61471735707245e-05, "loss": 0.0607, "step": 13560 }, { "grad_norm": 0.5727916955947876, "learning_rate": 9.613921363766592e-05, "loss": 0.0534, "step": 13570 }, { "grad_norm": 0.48917466402053833, "learning_rate": 9.613124582067763e-05, "loss": 0.0488, "step": 13580 }, { "grad_norm": 0.3412292003631592, "learning_rate": 9.612327012112112e-05, "loss": 0.0698, "step": 13590 }, { "grad_norm": 0.44028517603874207, "learning_rate": 9.611528654035921e-05, "loss": 0.077, "step": 13600 }, { "grad_norm": 0.3758545219898224, "learning_rate": 9.610729507975611e-05, "loss": 0.0491, "step": 13610 }, { "grad_norm": 0.4085526466369629, "learning_rate": 9.609929574067731e-05, "loss": 0.0564, "step": 13620 }, { "grad_norm": 0.4668867886066437, "learning_rate": 9.609128852448967e-05, "loss": 0.0483, "step": 13630 }, { "grad_norm": 0.43089261651039124, "learning_rate": 9.608327343256143e-05, "loss": 0.0519, "step": 13640 }, { "grad_norm": 0.49530526995658875, "learning_rate": 9.607525046626216e-05, "loss": 0.0573, "step": 13650 }, { "grad_norm": 0.577147364616394, "learning_rate": 9.606721962696272e-05, "loss": 0.0664, "step": 13660 }, { "grad_norm": 0.4671688973903656, "learning_rate": 9.60591809160354e-05, "loss": 0.0465, "step": 13670 }, { "grad_norm": 0.43772733211517334, "learning_rate": 9.605113433485378e-05, "loss": 0.0546, "step": 13680 }, { "grad_norm": 0.5903952121734619, "learning_rate": 9.604307988479279e-05, "loss": 0.0548, "step": 13690 }, { "grad_norm": 0.45376986265182495, "learning_rate": 9.603501756722876e-05, "loss": 0.0457, "step": 13700 }, { "grad_norm": 0.45929571986198425, "learning_rate": 9.602694738353927e-05, "loss": 0.0491, "step": 13710 }, { "grad_norm": 0.3843420445919037, "learning_rate": 9.601886933510331e-05, "loss": 0.0495, "step": 13720 }, { "grad_norm": 0.5027032494544983, "learning_rate": 9.60107834233012e-05, "loss": 0.0498, "step": 13730 }, { "grad_norm": 0.6337907314300537, "learning_rate": 9.60026896495146e-05, "loss": 0.0426, "step": 13740 }, { "grad_norm": 0.2714245319366455, "learning_rate": 9.599458801512652e-05, "loss": 0.0495, "step": 13750 }, { "grad_norm": 0.5184109807014465, "learning_rate": 9.598647852152129e-05, "loss": 0.0742, "step": 13760 }, { "grad_norm": 0.3258274495601654, "learning_rate": 9.597836117008462e-05, "loss": 0.0432, "step": 13770 }, { "grad_norm": 0.48920243978500366, "learning_rate": 9.597023596220356e-05, "loss": 0.0502, "step": 13780 }, { "grad_norm": 0.5243143439292908, "learning_rate": 9.596210289926643e-05, "loss": 0.0499, "step": 13790 }, { "grad_norm": 0.429219126701355, "learning_rate": 9.5953961982663e-05, "loss": 0.0445, "step": 13800 }, { "grad_norm": 0.48539865016937256, "learning_rate": 9.594581321378431e-05, "loss": 0.0638, "step": 13810 }, { "grad_norm": 0.5246344804763794, "learning_rate": 9.593765659402276e-05, "loss": 0.0532, "step": 13820 }, { "grad_norm": 0.3864213824272156, "learning_rate": 9.59294921247721e-05, "loss": 0.0561, "step": 13830 }, { "grad_norm": 0.4212910532951355, "learning_rate": 9.59213198074274e-05, "loss": 0.0541, "step": 13840 }, { "grad_norm": 0.4398423433303833, "learning_rate": 9.59131396433851e-05, "loss": 0.0442, "step": 13850 }, { "grad_norm": 0.3840423822402954, "learning_rate": 9.590495163404297e-05, "loss": 0.0479, "step": 13860 }, { "grad_norm": 0.41421452164649963, "learning_rate": 9.589675578080009e-05, "loss": 0.0587, "step": 13870 }, { "grad_norm": 0.4062683880329132, "learning_rate": 9.588855208505694e-05, "loss": 0.0416, "step": 13880 }, { "grad_norm": 0.5546085238456726, "learning_rate": 9.588034054821529e-05, "loss": 0.061, "step": 13890 }, { "grad_norm": 0.31342262029647827, "learning_rate": 9.587212117167826e-05, "loss": 0.0498, "step": 13900 }, { "grad_norm": 0.5205070972442627, "learning_rate": 9.586389395685033e-05, "loss": 0.0527, "step": 13910 }, { "grad_norm": 0.4851172864437103, "learning_rate": 9.585565890513733e-05, "loss": 0.0493, "step": 13920 }, { "grad_norm": 0.37505069375038147, "learning_rate": 9.584741601794636e-05, "loss": 0.0474, "step": 13930 }, { "grad_norm": 0.4620378613471985, "learning_rate": 9.58391652966859e-05, "loss": 0.0494, "step": 13940 }, { "grad_norm": 0.4466128945350647, "learning_rate": 9.583090674276583e-05, "loss": 0.0502, "step": 13950 }, { "grad_norm": 0.3884567320346832, "learning_rate": 9.582264035759726e-05, "loss": 0.0564, "step": 13960 }, { "grad_norm": 0.43840640783309937, "learning_rate": 9.58143661425927e-05, "loss": 0.0703, "step": 13970 }, { "grad_norm": 0.5855480432510376, "learning_rate": 9.580608409916601e-05, "loss": 0.0708, "step": 13980 }, { "grad_norm": 0.36561352014541626, "learning_rate": 9.579779422873233e-05, "loss": 0.0513, "step": 13990 }, { "grad_norm": 0.4923684298992157, "learning_rate": 9.578949653270819e-05, "loss": 0.0517, "step": 14000 }, { "grad_norm": 0.3992217183113098, "learning_rate": 9.578119101251144e-05, "loss": 0.0477, "step": 14010 }, { "grad_norm": 0.6274110078811646, "learning_rate": 9.577287766956127e-05, "loss": 0.0633, "step": 14020 }, { "grad_norm": 0.5647887587547302, "learning_rate": 9.57645565052782e-05, "loss": 0.0641, "step": 14030 }, { "grad_norm": 0.5389299392700195, "learning_rate": 9.575622752108407e-05, "loss": 0.0476, "step": 14040 }, { "grad_norm": 0.47353994846343994, "learning_rate": 9.57478907184021e-05, "loss": 0.0486, "step": 14050 }, { "grad_norm": 0.4572065472602844, "learning_rate": 9.573954609865681e-05, "loss": 0.0563, "step": 14060 }, { "grad_norm": 0.4644348621368408, "learning_rate": 9.573119366327408e-05, "loss": 0.0543, "step": 14070 }, { "grad_norm": 0.5742934346199036, "learning_rate": 9.57228334136811e-05, "loss": 0.0507, "step": 14080 }, { "grad_norm": 0.3623291850090027, "learning_rate": 9.571446535130641e-05, "loss": 0.0449, "step": 14090 }, { "grad_norm": 0.35291263461112976, "learning_rate": 9.570608947757988e-05, "loss": 0.0518, "step": 14100 }, { "grad_norm": 0.43009403347969055, "learning_rate": 9.569770579393274e-05, "loss": 0.0519, "step": 14110 }, { "grad_norm": 0.33192354440689087, "learning_rate": 9.56893143017975e-05, "loss": 0.0411, "step": 14120 }, { "grad_norm": 0.48797357082366943, "learning_rate": 9.568091500260806e-05, "loss": 0.0569, "step": 14130 }, { "grad_norm": 0.40111804008483887, "learning_rate": 9.567250789779961e-05, "loss": 0.0471, "step": 14140 }, { "grad_norm": 0.41295966506004333, "learning_rate": 9.566409298880872e-05, "loss": 0.046, "step": 14150 }, { "grad_norm": 0.7393625378608704, "learning_rate": 9.565567027707326e-05, "loss": 0.0675, "step": 14160 }, { "grad_norm": 0.4173564910888672, "learning_rate": 9.56472397640324e-05, "loss": 0.0512, "step": 14170 }, { "grad_norm": 0.48233288526535034, "learning_rate": 9.563880145112675e-05, "loss": 0.0526, "step": 14180 }, { "grad_norm": 0.43384087085723877, "learning_rate": 9.563035533979814e-05, "loss": 0.0521, "step": 14190 }, { "grad_norm": 0.8463438153266907, "learning_rate": 9.562190143148981e-05, "loss": 0.0439, "step": 14200 }, { "grad_norm": 0.43206965923309326, "learning_rate": 9.561343972764627e-05, "loss": 0.0524, "step": 14210 }, { "grad_norm": 0.39059677720069885, "learning_rate": 9.560497022971343e-05, "loss": 0.0582, "step": 14220 }, { "grad_norm": 0.49591299891471863, "learning_rate": 9.559649293913847e-05, "loss": 0.0582, "step": 14230 }, { "grad_norm": 0.5708838701248169, "learning_rate": 9.558800785736993e-05, "loss": 0.0486, "step": 14240 }, { "grad_norm": 0.530379056930542, "learning_rate": 9.557951498585767e-05, "loss": 0.0521, "step": 14250 }, { "grad_norm": 0.42820224165916443, "learning_rate": 9.557101432605293e-05, "loss": 0.0566, "step": 14260 }, { "grad_norm": 0.496303528547287, "learning_rate": 9.556250587940818e-05, "loss": 0.0664, "step": 14270 }, { "grad_norm": 0.5247111320495605, "learning_rate": 9.555398964737734e-05, "loss": 0.0507, "step": 14280 }, { "grad_norm": 0.3721764385700226, "learning_rate": 9.554546563141555e-05, "loss": 0.0483, "step": 14290 }, { "grad_norm": 0.41287747025489807, "learning_rate": 9.553693383297937e-05, "loss": 0.0605, "step": 14300 }, { "grad_norm": 0.60091632604599, "learning_rate": 9.552839425352663e-05, "loss": 0.0502, "step": 14310 }, { "grad_norm": 0.4947604537010193, "learning_rate": 9.551984689451652e-05, "loss": 0.0693, "step": 14320 }, { "grad_norm": 0.36028245091438293, "learning_rate": 9.551129175740953e-05, "loss": 0.0568, "step": 14330 }, { "grad_norm": 0.49672576785087585, "learning_rate": 9.550272884366754e-05, "loss": 0.059, "step": 14340 }, { "grad_norm": 0.5566672682762146, "learning_rate": 9.549415815475369e-05, "loss": 0.0578, "step": 14350 }, { "grad_norm": 0.5293188691139221, "learning_rate": 9.548557969213247e-05, "loss": 0.0561, "step": 14360 }, { "grad_norm": 0.5501118898391724, "learning_rate": 9.547699345726972e-05, "loss": 0.0607, "step": 14370 }, { "grad_norm": 0.39573800563812256, "learning_rate": 9.546839945163257e-05, "loss": 0.0521, "step": 14380 }, { "grad_norm": 0.4395396113395691, "learning_rate": 9.545979767668953e-05, "loss": 0.061, "step": 14390 }, { "grad_norm": 0.4503886103630066, "learning_rate": 9.54511881339104e-05, "loss": 0.0582, "step": 14400 }, { "grad_norm": 0.7075271606445312, "learning_rate": 9.54425708247663e-05, "loss": 0.0475, "step": 14410 }, { "grad_norm": 0.4079468846321106, "learning_rate": 9.543394575072972e-05, "loss": 0.0593, "step": 14420 }, { "grad_norm": 0.4637731909751892, "learning_rate": 9.542531291327441e-05, "loss": 0.0632, "step": 14430 }, { "grad_norm": 0.5930014252662659, "learning_rate": 9.541667231387552e-05, "loss": 0.0586, "step": 14440 }, { "grad_norm": 0.9561004638671875, "learning_rate": 9.540802395400949e-05, "loss": 0.0709, "step": 14450 }, { "grad_norm": 0.5417441725730896, "learning_rate": 9.539936783515406e-05, "loss": 0.0649, "step": 14460 }, { "grad_norm": 0.5083252787590027, "learning_rate": 9.539070395878835e-05, "loss": 0.0599, "step": 14470 }, { "grad_norm": 0.6383900046348572, "learning_rate": 9.538203232639277e-05, "loss": 0.0606, "step": 14480 }, { "grad_norm": 0.5272000432014465, "learning_rate": 9.537335293944907e-05, "loss": 0.0565, "step": 14490 }, { "grad_norm": 0.35628944635391235, "learning_rate": 9.536466579944032e-05, "loss": 0.0726, "step": 14500 }, { "grad_norm": 0.42774999141693115, "learning_rate": 9.535597090785091e-05, "loss": 0.0565, "step": 14510 }, { "grad_norm": 0.48928940296173096, "learning_rate": 9.534726826616656e-05, "loss": 0.0565, "step": 14520 }, { "grad_norm": 0.6565300226211548, "learning_rate": 9.53385578758743e-05, "loss": 0.0869, "step": 14530 }, { "grad_norm": 0.41470739245414734, "learning_rate": 9.532983973846252e-05, "loss": 0.0539, "step": 14540 }, { "grad_norm": 0.4351426362991333, "learning_rate": 9.53211138554209e-05, "loss": 0.0515, "step": 14550 }, { "grad_norm": 0.4370235502719879, "learning_rate": 9.531238022824047e-05, "loss": 0.055, "step": 14560 }, { "grad_norm": 0.42906826734542847, "learning_rate": 9.530363885841355e-05, "loss": 0.04, "step": 14570 }, { "grad_norm": 0.6274909973144531, "learning_rate": 9.52948897474338e-05, "loss": 0.0599, "step": 14580 }, { "grad_norm": 0.7572312951087952, "learning_rate": 9.528613289679622e-05, "loss": 0.073, "step": 14590 }, { "grad_norm": 0.6117717027664185, "learning_rate": 9.52773683079971e-05, "loss": 0.0584, "step": 14600 }, { "grad_norm": 0.3331352770328522, "learning_rate": 9.526859598253407e-05, "loss": 0.0489, "step": 14610 }, { "grad_norm": 0.4494798183441162, "learning_rate": 9.525981592190609e-05, "loss": 0.0454, "step": 14620 }, { "grad_norm": 0.4034304618835449, "learning_rate": 9.525102812761342e-05, "loss": 0.0551, "step": 14630 }, { "grad_norm": 0.5438587665557861, "learning_rate": 9.524223260115768e-05, "loss": 0.0416, "step": 14640 }, { "grad_norm": 0.4704503118991852, "learning_rate": 9.523342934404175e-05, "loss": 0.0502, "step": 14650 }, { "grad_norm": 0.4119521379470825, "learning_rate": 9.522461835776989e-05, "loss": 0.0452, "step": 14660 }, { "grad_norm": 0.4690583646297455, "learning_rate": 9.521579964384764e-05, "loss": 0.0509, "step": 14670 }, { "grad_norm": 0.5293126702308655, "learning_rate": 9.52069732037819e-05, "loss": 0.0515, "step": 14680 }, { "grad_norm": 0.438334584236145, "learning_rate": 9.519813903908083e-05, "loss": 0.0555, "step": 14690 }, { "grad_norm": 0.48403945565223694, "learning_rate": 9.5189297151254e-05, "loss": 0.0534, "step": 14700 }, { "grad_norm": 0.6420902013778687, "learning_rate": 9.518044754181218e-05, "loss": 0.0567, "step": 14710 }, { "grad_norm": 0.5119578242301941, "learning_rate": 9.51715902122676e-05, "loss": 0.0576, "step": 14720 }, { "grad_norm": 0.623162031173706, "learning_rate": 9.516272516413368e-05, "loss": 0.0564, "step": 14730 }, { "grad_norm": 0.5464646816253662, "learning_rate": 9.515385239892525e-05, "loss": 0.0587, "step": 14740 }, { "grad_norm": 0.5453888773918152, "learning_rate": 9.514497191815839e-05, "loss": 0.0482, "step": 14750 }, { "grad_norm": 0.4740324318408966, "learning_rate": 9.513608372335055e-05, "loss": 0.0547, "step": 14760 }, { "grad_norm": 0.47434958815574646, "learning_rate": 9.512718781602045e-05, "loss": 0.0506, "step": 14770 }, { "grad_norm": 0.5428524613380432, "learning_rate": 9.511828419768823e-05, "loss": 0.0617, "step": 14780 }, { "grad_norm": 0.4017029404640198, "learning_rate": 9.510937286987521e-05, "loss": 0.0456, "step": 14790 }, { "grad_norm": 0.5265960693359375, "learning_rate": 9.510045383410408e-05, "loss": 0.0572, "step": 14800 }, { "grad_norm": 0.4460238814353943, "learning_rate": 9.509152709189892e-05, "loss": 0.0524, "step": 14810 }, { "grad_norm": 0.4775349795818329, "learning_rate": 9.508259264478504e-05, "loss": 0.0611, "step": 14820 }, { "grad_norm": 0.4361923933029175, "learning_rate": 9.507365049428909e-05, "loss": 0.0707, "step": 14830 }, { "grad_norm": 0.7365840673446655, "learning_rate": 9.506470064193902e-05, "loss": 0.0673, "step": 14840 }, { "grad_norm": 0.3955390155315399, "learning_rate": 9.505574308926414e-05, "loss": 0.0631, "step": 14850 }, { "grad_norm": 0.42665934562683105, "learning_rate": 9.504677783779505e-05, "loss": 0.0462, "step": 14860 }, { "grad_norm": 0.28511151671409607, "learning_rate": 9.503780488906365e-05, "loss": 0.0445, "step": 14870 }, { "grad_norm": 0.47809723019599915, "learning_rate": 9.502882424460319e-05, "loss": 0.0542, "step": 14880 }, { "grad_norm": 0.6189029812812805, "learning_rate": 9.501983590594821e-05, "loss": 0.0558, "step": 14890 }, { "grad_norm": 0.4932357966899872, "learning_rate": 9.501083987463455e-05, "loss": 0.048, "step": 14900 }, { "grad_norm": 0.4261172413825989, "learning_rate": 9.500183615219942e-05, "loss": 0.0513, "step": 14910 }, { "grad_norm": 0.4014977514743805, "learning_rate": 9.49928247401813e-05, "loss": 0.0406, "step": 14920 }, { "grad_norm": 0.49482735991477966, "learning_rate": 9.498380564011997e-05, "loss": 0.044, "step": 14930 }, { "grad_norm": 0.41843917965888977, "learning_rate": 9.497477885355656e-05, "loss": 0.0584, "step": 14940 }, { "grad_norm": 0.5371150374412537, "learning_rate": 9.496574438203353e-05, "loss": 0.0626, "step": 14950 }, { "grad_norm": 0.40064018964767456, "learning_rate": 9.495670222709459e-05, "loss": 0.0523, "step": 14960 }, { "grad_norm": 0.4438404142856598, "learning_rate": 9.494765239028483e-05, "loss": 0.0494, "step": 14970 }, { "grad_norm": 0.42296475172042847, "learning_rate": 9.493859487315057e-05, "loss": 0.0538, "step": 14980 }, { "grad_norm": 0.4308440089225769, "learning_rate": 9.492952967723953e-05, "loss": 0.0477, "step": 14990 }, { "grad_norm": 0.46461907029151917, "learning_rate": 9.492045680410068e-05, "loss": 0.0569, "step": 15000 }, { "grad_norm": 0.3839108943939209, "learning_rate": 9.491137625528436e-05, "loss": 0.038, "step": 15010 }, { "grad_norm": 0.539222240447998, "learning_rate": 9.490228803234215e-05, "loss": 0.055, "step": 15020 }, { "grad_norm": 0.5233969688415527, "learning_rate": 9.489319213682701e-05, "loss": 0.0511, "step": 15030 }, { "grad_norm": 0.5275737643241882, "learning_rate": 9.488408857029316e-05, "loss": 0.0644, "step": 15040 }, { "grad_norm": 0.37501955032348633, "learning_rate": 9.487497733429616e-05, "loss": 0.0602, "step": 15050 }, { "grad_norm": 0.4312286078929901, "learning_rate": 9.486585843039286e-05, "loss": 0.0526, "step": 15060 }, { "grad_norm": 0.7532163858413696, "learning_rate": 9.485673186014143e-05, "loss": 0.0528, "step": 15070 }, { "grad_norm": 0.3869307041168213, "learning_rate": 9.484759762510137e-05, "loss": 0.0592, "step": 15080 }, { "grad_norm": 0.5284066796302795, "learning_rate": 9.483845572683346e-05, "loss": 0.0554, "step": 15090 }, { "grad_norm": 0.4637015461921692, "learning_rate": 9.48293061668998e-05, "loss": 0.0584, "step": 15100 }, { "grad_norm": 0.44787225127220154, "learning_rate": 9.48201489468638e-05, "loss": 0.0484, "step": 15110 }, { "grad_norm": 0.47783732414245605, "learning_rate": 9.481098406829016e-05, "loss": 0.0503, "step": 15120 }, { "grad_norm": 0.4272715747356415, "learning_rate": 9.480181153274495e-05, "loss": 0.0466, "step": 15130 }, { "grad_norm": 0.36535248160362244, "learning_rate": 9.479263134179548e-05, "loss": 0.0389, "step": 15140 }, { "grad_norm": 0.4102577567100525, "learning_rate": 9.478344349701039e-05, "loss": 0.0465, "step": 15150 }, { "grad_norm": 0.32607701420783997, "learning_rate": 9.477424799995964e-05, "loss": 0.0403, "step": 15160 }, { "grad_norm": 0.5599385499954224, "learning_rate": 9.476504485221448e-05, "loss": 0.0598, "step": 15170 }, { "grad_norm": 0.45458465814590454, "learning_rate": 9.475583405534748e-05, "loss": 0.0505, "step": 15180 }, { "grad_norm": 0.35848909616470337, "learning_rate": 9.474661561093251e-05, "loss": 0.0475, "step": 15190 }, { "grad_norm": 0.6347249746322632, "learning_rate": 9.473738952054478e-05, "loss": 0.0513, "step": 15200 }, { "grad_norm": 0.3891034424304962, "learning_rate": 9.472815578576073e-05, "loss": 0.0399, "step": 15210 }, { "grad_norm": 0.6256058216094971, "learning_rate": 9.471891440815817e-05, "loss": 0.0449, "step": 15220 }, { "grad_norm": 0.3689767122268677, "learning_rate": 9.470966538931621e-05, "loss": 0.0434, "step": 15230 }, { "grad_norm": 0.3840538561344147, "learning_rate": 9.470040873081525e-05, "loss": 0.038, "step": 15240 }, { "grad_norm": 0.5269218683242798, "learning_rate": 9.469114443423698e-05, "loss": 0.0556, "step": 15250 }, { "grad_norm": 0.4722138047218323, "learning_rate": 9.468187250116445e-05, "loss": 0.0509, "step": 15260 }, { "grad_norm": 0.36822509765625, "learning_rate": 9.467259293318197e-05, "loss": 0.0403, "step": 15270 }, { "grad_norm": 0.37308916449546814, "learning_rate": 9.466330573187514e-05, "loss": 0.0476, "step": 15280 }, { "grad_norm": 0.4482436776161194, "learning_rate": 9.46540108988309e-05, "loss": 0.0522, "step": 15290 }, { "grad_norm": 0.4083167016506195, "learning_rate": 9.46447084356375e-05, "loss": 0.0521, "step": 15300 }, { "grad_norm": 0.3571448028087616, "learning_rate": 9.463539834388447e-05, "loss": 0.0611, "step": 15310 }, { "grad_norm": 0.5266691446304321, "learning_rate": 9.462608062516263e-05, "loss": 0.0504, "step": 15320 }, { "grad_norm": 0.472105473279953, "learning_rate": 9.461675528106413e-05, "loss": 0.0408, "step": 15330 }, { "grad_norm": 0.3775545656681061, "learning_rate": 9.460742231318244e-05, "loss": 0.0516, "step": 15340 }, { "grad_norm": 0.6589304804801941, "learning_rate": 9.459808172311229e-05, "loss": 0.0555, "step": 15350 }, { "grad_norm": 0.4530782997608185, "learning_rate": 9.458873351244972e-05, "loss": 0.0457, "step": 15360 }, { "grad_norm": 0.37818413972854614, "learning_rate": 9.457937768279211e-05, "loss": 0.0454, "step": 15370 }, { "grad_norm": 0.4056207835674286, "learning_rate": 9.45700142357381e-05, "loss": 0.0475, "step": 15380 }, { "grad_norm": 0.6158747673034668, "learning_rate": 9.456064317288765e-05, "loss": 0.0422, "step": 15390 }, { "grad_norm": 0.4580637812614441, "learning_rate": 9.455126449584201e-05, "loss": 0.045, "step": 15400 }, { "grad_norm": 0.4400653839111328, "learning_rate": 9.454187820620375e-05, "loss": 0.0388, "step": 15410 }, { "grad_norm": 0.3231004774570465, "learning_rate": 9.453248430557673e-05, "loss": 0.0466, "step": 15420 }, { "grad_norm": 0.43058058619499207, "learning_rate": 9.452308279556611e-05, "loss": 0.0705, "step": 15430 }, { "grad_norm": 0.4986167252063751, "learning_rate": 9.451367367777835e-05, "loss": 0.0493, "step": 15440 }, { "grad_norm": 0.5389044880867004, "learning_rate": 9.450425695382122e-05, "loss": 0.062, "step": 15450 }, { "grad_norm": 0.45862776041030884, "learning_rate": 9.449483262530375e-05, "loss": 0.0504, "step": 15460 }, { "grad_norm": 0.6469208598136902, "learning_rate": 9.448540069383633e-05, "loss": 0.042, "step": 15470 }, { "grad_norm": 0.40831732749938965, "learning_rate": 9.447596116103061e-05, "loss": 0.047, "step": 15480 }, { "grad_norm": 0.48060137033462524, "learning_rate": 9.446651402849955e-05, "loss": 0.0563, "step": 15490 }, { "grad_norm": 0.40953686833381653, "learning_rate": 9.44570592978574e-05, "loss": 0.0508, "step": 15500 }, { "grad_norm": 0.34576600790023804, "learning_rate": 9.444759697071972e-05, "loss": 0.0512, "step": 15510 }, { "grad_norm": 0.3831237256526947, "learning_rate": 9.443812704870336e-05, "loss": 0.0413, "step": 15520 }, { "grad_norm": 0.3559820353984833, "learning_rate": 9.442864953342649e-05, "loss": 0.0418, "step": 15530 }, { "grad_norm": 0.39590704441070557, "learning_rate": 9.441916442650852e-05, "loss": 0.0485, "step": 15540 }, { "grad_norm": 0.48931270837783813, "learning_rate": 9.440967172957023e-05, "loss": 0.0444, "step": 15550 }, { "grad_norm": 0.44173529744148254, "learning_rate": 9.440017144423364e-05, "loss": 0.0615, "step": 15560 }, { "grad_norm": 0.3786730468273163, "learning_rate": 9.439066357212209e-05, "loss": 0.0512, "step": 15570 }, { "grad_norm": 0.4493815004825592, "learning_rate": 9.438114811486022e-05, "loss": 0.0611, "step": 15580 }, { "grad_norm": 0.37136900424957275, "learning_rate": 9.4371625074074e-05, "loss": 0.0442, "step": 15590 }, { "grad_norm": 0.4462362229824066, "learning_rate": 9.436209445139059e-05, "loss": 0.0482, "step": 15600 }, { "grad_norm": 0.387409508228302, "learning_rate": 9.435255624843855e-05, "loss": 0.0615, "step": 15610 }, { "grad_norm": 0.34305325150489807, "learning_rate": 9.43430104668477e-05, "loss": 0.0429, "step": 15620 }, { "grad_norm": 0.5299250483512878, "learning_rate": 9.433345710824914e-05, "loss": 0.0486, "step": 15630 }, { "grad_norm": 0.5074911117553711, "learning_rate": 9.432389617427529e-05, "loss": 0.0479, "step": 15640 }, { "grad_norm": 0.4061799645423889, "learning_rate": 9.431432766655984e-05, "loss": 0.0462, "step": 15650 }, { "grad_norm": 0.5320041179656982, "learning_rate": 9.430475158673778e-05, "loss": 0.0505, "step": 15660 }, { "grad_norm": 0.40959712862968445, "learning_rate": 9.429516793644542e-05, "loss": 0.0434, "step": 15670 }, { "grad_norm": 0.40376147627830505, "learning_rate": 9.428557671732034e-05, "loss": 0.0422, "step": 15680 }, { "grad_norm": 0.3952743709087372, "learning_rate": 9.42759779310014e-05, "loss": 0.0474, "step": 15690 }, { "grad_norm": 0.43282386660575867, "learning_rate": 9.426637157912879e-05, "loss": 0.0487, "step": 15700 }, { "grad_norm": 0.6386902928352356, "learning_rate": 9.425675766334397e-05, "loss": 0.0538, "step": 15710 }, { "grad_norm": 0.5251457691192627, "learning_rate": 9.424713618528968e-05, "loss": 0.0458, "step": 15720 }, { "grad_norm": 0.5966806411743164, "learning_rate": 9.423750714661e-05, "loss": 0.0536, "step": 15730 }, { "grad_norm": 0.45355111360549927, "learning_rate": 9.422787054895022e-05, "loss": 0.0483, "step": 15740 }, { "grad_norm": 0.43895798921585083, "learning_rate": 9.4218226393957e-05, "loss": 0.0476, "step": 15750 }, { "grad_norm": 0.46752867102622986, "learning_rate": 9.420857468327828e-05, "loss": 0.0508, "step": 15760 }, { "grad_norm": 0.45037683844566345, "learning_rate": 9.419891541856323e-05, "loss": 0.054, "step": 15770 }, { "grad_norm": 0.31222882866859436, "learning_rate": 9.41892486014624e-05, "loss": 0.0429, "step": 15780 }, { "grad_norm": 0.4638271629810333, "learning_rate": 9.417957423362756e-05, "loss": 0.0542, "step": 15790 }, { "grad_norm": 0.3989512324333191, "learning_rate": 9.416989231671178e-05, "loss": 0.0484, "step": 15800 }, { "grad_norm": 0.46569257974624634, "learning_rate": 9.416020285236946e-05, "loss": 0.0425, "step": 15810 }, { "grad_norm": 0.365618497133255, "learning_rate": 9.415050584225626e-05, "loss": 0.0389, "step": 15820 }, { "grad_norm": 0.5206946730613708, "learning_rate": 9.414080128802914e-05, "loss": 0.0466, "step": 15830 }, { "grad_norm": 0.46293941140174866, "learning_rate": 9.413108919134632e-05, "loss": 0.0471, "step": 15840 }, { "grad_norm": 0.46951019763946533, "learning_rate": 9.412136955386734e-05, "loss": 0.0609, "step": 15850 }, { "grad_norm": 0.31781265139579773, "learning_rate": 9.411164237725303e-05, "loss": 0.0487, "step": 15860 }, { "grad_norm": 0.47759467363357544, "learning_rate": 9.41019076631655e-05, "loss": 0.0519, "step": 15870 }, { "grad_norm": 0.43992340564727783, "learning_rate": 9.409216541326815e-05, "loss": 0.0617, "step": 15880 }, { "grad_norm": 0.4674963057041168, "learning_rate": 9.408241562922564e-05, "loss": 0.0603, "step": 15890 }, { "grad_norm": 0.36306291818618774, "learning_rate": 9.407265831270395e-05, "loss": 0.0357, "step": 15900 }, { "grad_norm": 0.4497736096382141, "learning_rate": 9.406289346537035e-05, "loss": 0.0512, "step": 15910 }, { "grad_norm": 0.4525422155857086, "learning_rate": 9.405312108889339e-05, "loss": 0.0512, "step": 15920 }, { "grad_norm": 0.47964486479759216, "learning_rate": 9.404334118494288e-05, "loss": 0.0559, "step": 15930 }, { "grad_norm": 0.35358697175979614, "learning_rate": 9.403355375518995e-05, "loss": 0.0462, "step": 15940 }, { "grad_norm": 0.478379487991333, "learning_rate": 9.4023758801307e-05, "loss": 0.045, "step": 15950 }, { "grad_norm": 0.288540780544281, "learning_rate": 9.401395632496774e-05, "loss": 0.042, "step": 15960 }, { "grad_norm": 0.446170449256897, "learning_rate": 9.400414632784711e-05, "loss": 0.0492, "step": 15970 }, { "grad_norm": 0.44895678758621216, "learning_rate": 9.39943288116214e-05, "loss": 0.0394, "step": 15980 }, { "grad_norm": 0.3828069269657135, "learning_rate": 9.398450377796815e-05, "loss": 0.0482, "step": 15990 }, { "grad_norm": 0.40511149168014526, "learning_rate": 9.397467122856616e-05, "loss": 0.0471, "step": 16000 }, { "grad_norm": 0.3311868906021118, "learning_rate": 9.396483116509558e-05, "loss": 0.0469, "step": 16010 }, { "grad_norm": 0.30421048402786255, "learning_rate": 9.39549835892378e-05, "loss": 0.0345, "step": 16020 }, { "grad_norm": 0.3245353102684021, "learning_rate": 9.39451285026755e-05, "loss": 0.0399, "step": 16030 }, { "grad_norm": 0.561484158039093, "learning_rate": 9.393526590709262e-05, "loss": 0.0442, "step": 16040 }, { "grad_norm": 0.4332939684391022, "learning_rate": 9.392539580417444e-05, "loss": 0.0504, "step": 16050 }, { "grad_norm": 0.6420621275901794, "learning_rate": 9.391551819560747e-05, "loss": 0.0538, "step": 16060 }, { "grad_norm": 0.4303891062736511, "learning_rate": 9.390563308307955e-05, "loss": 0.0497, "step": 16070 }, { "grad_norm": 0.4296530783176422, "learning_rate": 9.389574046827974e-05, "loss": 0.0536, "step": 16080 }, { "grad_norm": 0.5129997730255127, "learning_rate": 9.388584035289845e-05, "loss": 0.0535, "step": 16090 }, { "grad_norm": 0.4097578823566437, "learning_rate": 9.387593273862732e-05, "loss": 0.0528, "step": 16100 }, { "grad_norm": 0.41226890683174133, "learning_rate": 9.386601762715929e-05, "loss": 0.0463, "step": 16110 }, { "grad_norm": 0.4529740512371063, "learning_rate": 9.38560950201886e-05, "loss": 0.0529, "step": 16120 }, { "grad_norm": 0.44401898980140686, "learning_rate": 9.384616491941071e-05, "loss": 0.0458, "step": 16130 }, { "grad_norm": 0.516925573348999, "learning_rate": 9.383622732652245e-05, "loss": 0.0403, "step": 16140 }, { "grad_norm": 0.4691736698150635, "learning_rate": 9.382628224322187e-05, "loss": 0.0504, "step": 16150 }, { "grad_norm": 0.42201119661331177, "learning_rate": 9.381632967120829e-05, "loss": 0.0515, "step": 16160 }, { "grad_norm": 0.4651547074317932, "learning_rate": 9.380636961218235e-05, "loss": 0.0484, "step": 16170 }, { "grad_norm": 0.3721025884151459, "learning_rate": 9.379640206784597e-05, "loss": 0.0389, "step": 16180 }, { "grad_norm": 0.4671632945537567, "learning_rate": 9.378642703990229e-05, "loss": 0.0418, "step": 16190 }, { "grad_norm": 0.3737826943397522, "learning_rate": 9.37764445300558e-05, "loss": 0.0365, "step": 16200 }, { "grad_norm": 0.3959413766860962, "learning_rate": 9.376645454001222e-05, "loss": 0.0549, "step": 16210 }, { "grad_norm": 0.3806593120098114, "learning_rate": 9.375645707147858e-05, "loss": 0.0435, "step": 16220 }, { "grad_norm": 0.3708620071411133, "learning_rate": 9.374645212616316e-05, "loss": 0.0634, "step": 16230 }, { "grad_norm": 0.5943989157676697, "learning_rate": 9.373643970577555e-05, "loss": 0.0565, "step": 16240 }, { "grad_norm": 0.48789018392562866, "learning_rate": 9.372641981202659e-05, "loss": 0.0518, "step": 16250 }, { "grad_norm": 0.5299814939498901, "learning_rate": 9.37163924466284e-05, "loss": 0.053, "step": 16260 }, { "grad_norm": 0.34559929370880127, "learning_rate": 9.370635761129438e-05, "loss": 0.0466, "step": 16270 }, { "grad_norm": 0.458090603351593, "learning_rate": 9.36963153077392e-05, "loss": 0.0539, "step": 16280 }, { "grad_norm": 0.48743292689323425, "learning_rate": 9.368626553767888e-05, "loss": 0.0488, "step": 16290 }, { "grad_norm": 0.5373132228851318, "learning_rate": 9.367620830283057e-05, "loss": 0.0416, "step": 16300 }, { "grad_norm": 0.6262367367744446, "learning_rate": 9.366614360491281e-05, "loss": 0.0562, "step": 16310 }, { "grad_norm": 0.3529707193374634, "learning_rate": 9.365607144564539e-05, "loss": 0.0463, "step": 16320 }, { "grad_norm": 0.534331202507019, "learning_rate": 9.364599182674934e-05, "loss": 0.0633, "step": 16330 }, { "grad_norm": 0.4287354648113251, "learning_rate": 9.3635904749947e-05, "loss": 0.0525, "step": 16340 }, { "grad_norm": 0.48483768105506897, "learning_rate": 9.362581021696202e-05, "loss": 0.0398, "step": 16350 }, { "grad_norm": 0.4294423758983612, "learning_rate": 9.361570822951921e-05, "loss": 0.0488, "step": 16360 }, { "grad_norm": 0.4927031099796295, "learning_rate": 9.360559878934476e-05, "loss": 0.0434, "step": 16370 }, { "grad_norm": 0.31325793266296387, "learning_rate": 9.359548189816611e-05, "loss": 0.0391, "step": 16380 }, { "grad_norm": 0.5636285543441772, "learning_rate": 9.358535755771193e-05, "loss": 0.0559, "step": 16390 }, { "grad_norm": 0.47265592217445374, "learning_rate": 9.357522576971221e-05, "loss": 0.053, "step": 16400 }, { "grad_norm": 0.3956400156021118, "learning_rate": 9.356508653589819e-05, "loss": 0.0473, "step": 16410 }, { "grad_norm": 0.4368806779384613, "learning_rate": 9.355493985800237e-05, "loss": 0.042, "step": 16420 }, { "grad_norm": 0.6074290871620178, "learning_rate": 9.354478573775857e-05, "loss": 0.0509, "step": 16430 }, { "grad_norm": 0.42880815267562866, "learning_rate": 9.353462417690186e-05, "loss": 0.0476, "step": 16440 }, { "grad_norm": 0.4252299666404724, "learning_rate": 9.352445517716853e-05, "loss": 0.0387, "step": 16450 }, { "grad_norm": 0.38941437005996704, "learning_rate": 9.351427874029621e-05, "loss": 0.0569, "step": 16460 }, { "grad_norm": 0.4049396514892578, "learning_rate": 9.350409486802379e-05, "loss": 0.0459, "step": 16470 }, { "grad_norm": 0.4512145221233368, "learning_rate": 9.349390356209138e-05, "loss": 0.0488, "step": 16480 }, { "grad_norm": 0.3248178958892822, "learning_rate": 9.348370482424042e-05, "loss": 0.0496, "step": 16490 }, { "grad_norm": 0.5674871206283569, "learning_rate": 9.347349865621357e-05, "loss": 0.0425, "step": 16500 }, { "grad_norm": 0.3634752035140991, "learning_rate": 9.346328505975481e-05, "loss": 0.0422, "step": 16510 }, { "grad_norm": 0.4024083614349365, "learning_rate": 9.345306403660936e-05, "loss": 0.0515, "step": 16520 }, { "grad_norm": 0.5311237573623657, "learning_rate": 9.344283558852371e-05, "loss": 0.0439, "step": 16530 }, { "grad_norm": 0.5012393593788147, "learning_rate": 9.343259971724563e-05, "loss": 0.0498, "step": 16540 }, { "grad_norm": 0.5681824088096619, "learning_rate": 9.342235642452413e-05, "loss": 0.053, "step": 16550 }, { "grad_norm": 0.48208552598953247, "learning_rate": 9.341210571210954e-05, "loss": 0.0598, "step": 16560 }, { "grad_norm": 0.44742733240127563, "learning_rate": 9.340184758175338e-05, "loss": 0.0522, "step": 16570 }, { "grad_norm": 0.43088310956954956, "learning_rate": 9.339158203520854e-05, "loss": 0.0567, "step": 16580 }, { "grad_norm": 0.3995932936668396, "learning_rate": 9.338130907422908e-05, "loss": 0.0406, "step": 16590 }, { "grad_norm": 0.3199278712272644, "learning_rate": 9.337102870057037e-05, "loss": 0.0412, "step": 16600 }, { "grad_norm": 0.4373507499694824, "learning_rate": 9.336074091598907e-05, "loss": 0.0474, "step": 16610 }, { "grad_norm": 0.36832576990127563, "learning_rate": 9.335044572224306e-05, "loss": 0.0356, "step": 16620 }, { "grad_norm": 0.4929790198802948, "learning_rate": 9.334014312109151e-05, "loss": 0.0439, "step": 16630 }, { "grad_norm": 0.46662673354148865, "learning_rate": 9.332983311429486e-05, "loss": 0.0496, "step": 16640 }, { "grad_norm": 0.44499096274375916, "learning_rate": 9.33195157036148e-05, "loss": 0.05, "step": 16650 }, { "grad_norm": 0.538060188293457, "learning_rate": 9.330919089081432e-05, "loss": 0.0466, "step": 16660 }, { "grad_norm": 0.39456620812416077, "learning_rate": 9.32988586776576e-05, "loss": 0.0666, "step": 16670 }, { "grad_norm": 0.6637905836105347, "learning_rate": 9.328851906591016e-05, "loss": 0.0511, "step": 16680 }, { "grad_norm": 0.517956554889679, "learning_rate": 9.327817205733875e-05, "loss": 0.0509, "step": 16690 }, { "grad_norm": 0.45638930797576904, "learning_rate": 9.326781765371142e-05, "loss": 0.0504, "step": 16700 }, { "grad_norm": 0.4163559675216675, "learning_rate": 9.325745585679741e-05, "loss": 0.0585, "step": 16710 }, { "grad_norm": 0.4431612193584442, "learning_rate": 9.32470866683673e-05, "loss": 0.065, "step": 16720 }, { "grad_norm": 0.3964182734489441, "learning_rate": 9.323671009019288e-05, "loss": 0.0677, "step": 16730 }, { "grad_norm": 0.3356071412563324, "learning_rate": 9.322632612404725e-05, "loss": 0.0462, "step": 16740 }, { "grad_norm": 0.45155104994773865, "learning_rate": 9.321593477170471e-05, "loss": 0.0518, "step": 16750 }, { "grad_norm": 0.339783638715744, "learning_rate": 9.320553603494088e-05, "loss": 0.0402, "step": 16760 }, { "grad_norm": 0.38203462958335876, "learning_rate": 9.319512991553261e-05, "loss": 0.0455, "step": 16770 }, { "grad_norm": 0.3772253096103668, "learning_rate": 9.318471641525803e-05, "loss": 0.0437, "step": 16780 }, { "grad_norm": 0.368109792470932, "learning_rate": 9.317429553589652e-05, "loss": 0.0411, "step": 16790 }, { "grad_norm": 0.40550485253334045, "learning_rate": 9.316386727922873e-05, "loss": 0.0419, "step": 16800 }, { "grad_norm": 0.41872063279151917, "learning_rate": 9.315343164703656e-05, "loss": 0.0372, "step": 16810 }, { "grad_norm": 0.5546005964279175, "learning_rate": 9.314298864110316e-05, "loss": 0.0419, "step": 16820 }, { "grad_norm": 0.29309967160224915, "learning_rate": 9.313253826321295e-05, "loss": 0.0401, "step": 16830 }, { "grad_norm": 0.5213341116905212, "learning_rate": 9.312208051515165e-05, "loss": 0.0387, "step": 16840 }, { "grad_norm": 0.40111294388771057, "learning_rate": 9.311161539870618e-05, "loss": 0.0399, "step": 16850 }, { "grad_norm": 0.46857964992523193, "learning_rate": 9.310114291566474e-05, "loss": 0.0486, "step": 16860 }, { "grad_norm": 0.5463732481002808, "learning_rate": 9.309066306781679e-05, "loss": 0.0414, "step": 16870 }, { "grad_norm": 0.4116292893886566, "learning_rate": 9.308017585695306e-05, "loss": 0.0402, "step": 16880 }, { "grad_norm": 0.473425954580307, "learning_rate": 9.306968128486552e-05, "loss": 0.047, "step": 16890 }, { "grad_norm": 0.45104503631591797, "learning_rate": 9.30591793533474e-05, "loss": 0.0412, "step": 16900 }, { "grad_norm": 0.40683287382125854, "learning_rate": 9.304867006419321e-05, "loss": 0.0383, "step": 16910 }, { "grad_norm": 0.3842101991176605, "learning_rate": 9.303815341919868e-05, "loss": 0.0494, "step": 16920 }, { "grad_norm": 0.5758320689201355, "learning_rate": 9.302762942016084e-05, "loss": 0.046, "step": 16930 }, { "grad_norm": 0.3974977135658264, "learning_rate": 9.301709806887792e-05, "loss": 0.0458, "step": 16940 }, { "grad_norm": 0.3835126459598541, "learning_rate": 9.300655936714948e-05, "loss": 0.0591, "step": 16950 }, { "grad_norm": 0.40926671028137207, "learning_rate": 9.299601331677627e-05, "loss": 0.0449, "step": 16960 }, { "grad_norm": 0.43176236748695374, "learning_rate": 9.298545991956033e-05, "loss": 0.0467, "step": 16970 }, { "grad_norm": 0.38555026054382324, "learning_rate": 9.297489917730493e-05, "loss": 0.0412, "step": 16980 }, { "grad_norm": 0.5824466347694397, "learning_rate": 9.296433109181464e-05, "loss": 0.0513, "step": 16990 }, { "grad_norm": 0.6476307511329651, "learning_rate": 9.295375566489523e-05, "loss": 0.0436, "step": 17000 }, { "grad_norm": 0.5593743920326233, "learning_rate": 9.294317289835379e-05, "loss": 0.0597, "step": 17010 }, { "grad_norm": 0.4067201018333435, "learning_rate": 9.293258279399859e-05, "loss": 0.0584, "step": 17020 }, { "grad_norm": 0.34556689858436584, "learning_rate": 9.292198535363919e-05, "loss": 0.0422, "step": 17030 }, { "grad_norm": 0.47261643409729004, "learning_rate": 9.291138057908641e-05, "loss": 0.0406, "step": 17040 }, { "grad_norm": 0.48458901047706604, "learning_rate": 9.290076847215234e-05, "loss": 0.0463, "step": 17050 }, { "grad_norm": 0.6939793229103088, "learning_rate": 9.289014903465025e-05, "loss": 0.0516, "step": 17060 }, { "grad_norm": 0.40004533529281616, "learning_rate": 9.287952226839475e-05, "loss": 0.0414, "step": 17070 }, { "grad_norm": 0.4979913830757141, "learning_rate": 9.286888817520164e-05, "loss": 0.0461, "step": 17080 }, { "grad_norm": 0.34155726432800293, "learning_rate": 9.285824675688803e-05, "loss": 0.0442, "step": 17090 }, { "grad_norm": 0.38257697224617004, "learning_rate": 9.28475980152722e-05, "loss": 0.0466, "step": 17100 }, { "grad_norm": 0.4039163589477539, "learning_rate": 9.283694195217379e-05, "loss": 0.0482, "step": 17110 }, { "grad_norm": 0.36890488862991333, "learning_rate": 9.282627856941356e-05, "loss": 0.045, "step": 17120 }, { "grad_norm": 0.5212796926498413, "learning_rate": 9.281560786881363e-05, "loss": 0.039, "step": 17130 }, { "grad_norm": 0.32232847809791565, "learning_rate": 9.280492985219733e-05, "loss": 0.0408, "step": 17140 }, { "grad_norm": 0.33099398016929626, "learning_rate": 9.279424452138924e-05, "loss": 0.043, "step": 17150 }, { "grad_norm": 0.3555777370929718, "learning_rate": 9.278355187821517e-05, "loss": 0.0479, "step": 17160 }, { "grad_norm": 0.3373827338218689, "learning_rate": 9.277285192450224e-05, "loss": 0.0452, "step": 17170 }, { "grad_norm": 0.43394041061401367, "learning_rate": 9.276214466207875e-05, "loss": 0.0389, "step": 17180 }, { "grad_norm": 0.5010589957237244, "learning_rate": 9.275143009277427e-05, "loss": 0.0467, "step": 17190 }, { "grad_norm": 0.37650740146636963, "learning_rate": 9.274070821841964e-05, "loss": 0.0512, "step": 17200 }, { "grad_norm": 0.480754554271698, "learning_rate": 9.272997904084696e-05, "loss": 0.0503, "step": 17210 }, { "grad_norm": 0.49090877175331116, "learning_rate": 9.271924256188951e-05, "loss": 0.0588, "step": 17220 }, { "grad_norm": 0.43498146533966064, "learning_rate": 9.270849878338189e-05, "loss": 0.0558, "step": 17230 }, { "grad_norm": 0.43358415365219116, "learning_rate": 9.269774770715991e-05, "loss": 0.0476, "step": 17240 }, { "grad_norm": 0.4738347828388214, "learning_rate": 9.268698933506061e-05, "loss": 0.0672, "step": 17250 }, { "grad_norm": 0.3695964813232422, "learning_rate": 9.267622366892235e-05, "loss": 0.0496, "step": 17260 }, { "grad_norm": 0.3601413369178772, "learning_rate": 9.266545071058465e-05, "loss": 0.0428, "step": 17270 }, { "grad_norm": 0.37929317355155945, "learning_rate": 9.265467046188833e-05, "loss": 0.0443, "step": 17280 }, { "grad_norm": 0.42073291540145874, "learning_rate": 9.264388292467543e-05, "loss": 0.0437, "step": 17290 }, { "grad_norm": 0.39321351051330566, "learning_rate": 9.263308810078926e-05, "loss": 0.0405, "step": 17300 }, { "grad_norm": 0.319601446390152, "learning_rate": 9.262228599207434e-05, "loss": 0.0462, "step": 17310 }, { "grad_norm": 0.41791313886642456, "learning_rate": 9.261147660037647e-05, "loss": 0.0402, "step": 17320 }, { "grad_norm": 0.6572993397712708, "learning_rate": 9.26006599275427e-05, "loss": 0.0406, "step": 17330 }, { "grad_norm": 0.40072399377822876, "learning_rate": 9.258983597542124e-05, "loss": 0.0516, "step": 17340 }, { "grad_norm": 0.30850157141685486, "learning_rate": 9.257900474586167e-05, "loss": 0.0434, "step": 17350 }, { "grad_norm": 0.31466779112815857, "learning_rate": 9.256816624071471e-05, "loss": 0.051, "step": 17360 }, { "grad_norm": 0.41319218277931213, "learning_rate": 9.25573204618324e-05, "loss": 0.0388, "step": 17370 }, { "grad_norm": 0.4556146562099457, "learning_rate": 9.254646741106796e-05, "loss": 0.0429, "step": 17380 }, { "grad_norm": 0.5635148286819458, "learning_rate": 9.253560709027589e-05, "loss": 0.0436, "step": 17390 }, { "grad_norm": 0.41354531049728394, "learning_rate": 9.252473950131192e-05, "loss": 0.0472, "step": 17400 }, { "grad_norm": 0.48321235179901123, "learning_rate": 9.251386464603302e-05, "loss": 0.0378, "step": 17410 }, { "grad_norm": 0.435865193605423, "learning_rate": 9.250298252629741e-05, "loss": 0.0356, "step": 17420 }, { "grad_norm": 0.5960149168968201, "learning_rate": 9.249209314396454e-05, "loss": 0.0406, "step": 17430 }, { "grad_norm": 0.5934005379676819, "learning_rate": 9.248119650089513e-05, "loss": 0.0483, "step": 17440 }, { "grad_norm": 0.4891586899757385, "learning_rate": 9.247029259895108e-05, "loss": 0.0505, "step": 17450 }, { "grad_norm": 0.37167564034461975, "learning_rate": 9.24593814399956e-05, "loss": 0.0429, "step": 17460 }, { "grad_norm": 0.3033064305782318, "learning_rate": 9.244846302589309e-05, "loss": 0.0392, "step": 17470 }, { "grad_norm": 0.39480623602867126, "learning_rate": 9.243753735850923e-05, "loss": 0.0429, "step": 17480 }, { "grad_norm": 0.3331052362918854, "learning_rate": 9.24266044397109e-05, "loss": 0.0396, "step": 17490 }, { "grad_norm": 0.30568137764930725, "learning_rate": 9.241566427136624e-05, "loss": 0.0354, "step": 17500 }, { "grad_norm": 0.45848363637924194, "learning_rate": 9.240471685534463e-05, "loss": 0.0441, "step": 17510 }, { "grad_norm": 0.4047095477581024, "learning_rate": 9.239376219351667e-05, "loss": 0.042, "step": 17520 }, { "grad_norm": 0.3905563950538635, "learning_rate": 9.238280028775425e-05, "loss": 0.0393, "step": 17530 }, { "grad_norm": 0.46636858582496643, "learning_rate": 9.237183113993041e-05, "loss": 0.0511, "step": 17540 }, { "grad_norm": 0.46192434430122375, "learning_rate": 9.236085475191952e-05, "loss": 0.0399, "step": 17550 }, { "grad_norm": 0.38628825545310974, "learning_rate": 9.234987112559709e-05, "loss": 0.0519, "step": 17560 }, { "grad_norm": 0.6090779900550842, "learning_rate": 9.233888026283999e-05, "loss": 0.053, "step": 17570 }, { "grad_norm": 0.4099520742893219, "learning_rate": 9.232788216552619e-05, "loss": 0.0449, "step": 17580 }, { "grad_norm": 0.3976421356201172, "learning_rate": 9.231687683553502e-05, "loss": 0.0455, "step": 17590 }, { "grad_norm": 0.36644428968429565, "learning_rate": 9.230586427474698e-05, "loss": 0.0613, "step": 17600 }, { "grad_norm": 0.42456817626953125, "learning_rate": 9.229484448504379e-05, "loss": 0.0447, "step": 17610 }, { "grad_norm": 0.35498887300491333, "learning_rate": 9.228381746830843e-05, "loss": 0.0449, "step": 17620 }, { "grad_norm": 0.31657975912094116, "learning_rate": 9.227278322642514e-05, "loss": 0.0373, "step": 17630 }, { "grad_norm": 0.47756391763687134, "learning_rate": 9.226174176127937e-05, "loss": 0.0446, "step": 17640 }, { "grad_norm": 0.37552252411842346, "learning_rate": 9.22506930747578e-05, "loss": 0.041, "step": 17650 }, { "grad_norm": 0.5555407404899597, "learning_rate": 9.223963716874831e-05, "loss": 0.0576, "step": 17660 }, { "grad_norm": 0.46187809109687805, "learning_rate": 9.222857404514012e-05, "loss": 0.0488, "step": 17670 }, { "grad_norm": 0.44297653436660767, "learning_rate": 9.221750370582355e-05, "loss": 0.0433, "step": 17680 }, { "grad_norm": 0.43798935413360596, "learning_rate": 9.220642615269028e-05, "loss": 0.0382, "step": 17690 }, { "grad_norm": 0.29576393961906433, "learning_rate": 9.219534138763311e-05, "loss": 0.048, "step": 17700 }, { "grad_norm": 0.497376024723053, "learning_rate": 9.218424941254613e-05, "loss": 0.0435, "step": 17710 }, { "grad_norm": 0.4338616728782654, "learning_rate": 9.217315022932468e-05, "loss": 0.0413, "step": 17720 }, { "grad_norm": 0.2712116241455078, "learning_rate": 9.216204383986528e-05, "loss": 0.0369, "step": 17730 }, { "grad_norm": 0.45208197832107544, "learning_rate": 9.215093024606574e-05, "loss": 0.0574, "step": 17740 }, { "grad_norm": 0.5954285860061646, "learning_rate": 9.213980944982506e-05, "loss": 0.0588, "step": 17750 }, { "grad_norm": 0.5350531339645386, "learning_rate": 9.212868145304346e-05, "loss": 0.0607, "step": 17760 }, { "grad_norm": 0.4551539719104767, "learning_rate": 9.211754625762241e-05, "loss": 0.0369, "step": 17770 }, { "grad_norm": 0.4971333146095276, "learning_rate": 9.210640386546463e-05, "loss": 0.051, "step": 17780 }, { "grad_norm": 0.3407624065876007, "learning_rate": 9.209525427847405e-05, "loss": 0.0345, "step": 17790 }, { "grad_norm": 0.3766728341579437, "learning_rate": 9.208409749855583e-05, "loss": 0.0397, "step": 17800 }, { "grad_norm": 0.3608137369155884, "learning_rate": 9.207293352761633e-05, "loss": 0.0454, "step": 17810 }, { "grad_norm": 0.4906247854232788, "learning_rate": 9.206176236756319e-05, "loss": 0.0462, "step": 17820 }, { "grad_norm": 0.2690950632095337, "learning_rate": 9.205058402030525e-05, "loss": 0.042, "step": 17830 }, { "grad_norm": 0.5126135945320129, "learning_rate": 9.203939848775259e-05, "loss": 0.0476, "step": 17840 }, { "grad_norm": 0.42721500992774963, "learning_rate": 9.202820577181652e-05, "loss": 0.043, "step": 17850 }, { "grad_norm": 0.6185064315795898, "learning_rate": 9.201700587440953e-05, "loss": 0.056, "step": 17860 }, { "grad_norm": 0.3295431137084961, "learning_rate": 9.200579879744544e-05, "loss": 0.0493, "step": 17870 }, { "grad_norm": 0.8626433610916138, "learning_rate": 9.199458454283918e-05, "loss": 0.0473, "step": 17880 }, { "grad_norm": 0.38440585136413574, "learning_rate": 9.198336311250697e-05, "loss": 0.0453, "step": 17890 }, { "grad_norm": 0.5920936465263367, "learning_rate": 9.197213450836626e-05, "loss": 0.0499, "step": 17900 }, { "grad_norm": 0.953285276889801, "learning_rate": 9.19608987323357e-05, "loss": 0.0603, "step": 17910 }, { "grad_norm": 0.5511460304260254, "learning_rate": 9.194965578633517e-05, "loss": 0.0607, "step": 17920 }, { "grad_norm": 0.4082295596599579, "learning_rate": 9.193840567228582e-05, "loss": 0.0495, "step": 17930 }, { "grad_norm": 0.481758713722229, "learning_rate": 9.192714839210994e-05, "loss": 0.0465, "step": 17940 }, { "grad_norm": 0.3927723467350006, "learning_rate": 9.19158839477311e-05, "loss": 0.0473, "step": 17950 }, { "grad_norm": 0.49490416049957275, "learning_rate": 9.190461234107411e-05, "loss": 0.0479, "step": 17960 }, { "grad_norm": 0.3790321350097656, "learning_rate": 9.189333357406496e-05, "loss": 0.0368, "step": 17970 }, { "grad_norm": 0.31021878123283386, "learning_rate": 9.188204764863089e-05, "loss": 0.0529, "step": 17980 }, { "grad_norm": 0.5359218716621399, "learning_rate": 9.187075456670033e-05, "loss": 0.0382, "step": 17990 }, { "grad_norm": 0.3369406759738922, "learning_rate": 9.1859454330203e-05, "loss": 0.0531, "step": 18000 }, { "grad_norm": 0.35699978470802307, "learning_rate": 9.18481469410698e-05, "loss": 0.0563, "step": 18010 }, { "grad_norm": 0.3600858151912689, "learning_rate": 9.183683240123281e-05, "loss": 0.0495, "step": 18020 }, { "grad_norm": 0.4915299117565155, "learning_rate": 9.182551071262541e-05, "loss": 0.0482, "step": 18030 }, { "grad_norm": 0.47027385234832764, "learning_rate": 9.181418187718218e-05, "loss": 0.0521, "step": 18040 }, { "grad_norm": 0.5492985844612122, "learning_rate": 9.180284589683888e-05, "loss": 0.0425, "step": 18050 }, { "grad_norm": 0.44785815477371216, "learning_rate": 9.17915027735325e-05, "loss": 0.0388, "step": 18060 }, { "grad_norm": 0.34930598735809326, "learning_rate": 9.178015250920133e-05, "loss": 0.0478, "step": 18070 }, { "grad_norm": 0.327696830034256, "learning_rate": 9.176879510578477e-05, "loss": 0.036, "step": 18080 }, { "grad_norm": 0.5707911252975464, "learning_rate": 9.17574305652235e-05, "loss": 0.0427, "step": 18090 }, { "grad_norm": 0.4113367199897766, "learning_rate": 9.174605888945942e-05, "loss": 0.0519, "step": 18100 }, { "grad_norm": 0.5250842571258545, "learning_rate": 9.173468008043564e-05, "loss": 0.0502, "step": 18110 }, { "grad_norm": 0.4666343927383423, "learning_rate": 9.172329414009648e-05, "loss": 0.0462, "step": 18120 }, { "grad_norm": 0.4413873553276062, "learning_rate": 9.171190107038747e-05, "loss": 0.0434, "step": 18130 }, { "grad_norm": 0.34981393814086914, "learning_rate": 9.170050087325541e-05, "loss": 0.0412, "step": 18140 }, { "grad_norm": 0.4779450595378876, "learning_rate": 9.168909355064824e-05, "loss": 0.0527, "step": 18150 }, { "grad_norm": 0.5162726044654846, "learning_rate": 9.167767910451519e-05, "loss": 0.0455, "step": 18160 }, { "grad_norm": 0.3284902274608612, "learning_rate": 9.166625753680669e-05, "loss": 0.0425, "step": 18170 }, { "grad_norm": 0.527839183807373, "learning_rate": 9.165482884947431e-05, "loss": 0.0392, "step": 18180 }, { "grad_norm": 0.5702625513076782, "learning_rate": 9.164339304447098e-05, "loss": 0.0463, "step": 18190 }, { "grad_norm": 0.3698277175426483, "learning_rate": 9.163195012375072e-05, "loss": 0.0331, "step": 18200 }, { "grad_norm": 0.4063204824924469, "learning_rate": 9.16205000892688e-05, "loss": 0.0438, "step": 18210 }, { "grad_norm": 0.5041989684104919, "learning_rate": 9.160904294298175e-05, "loss": 0.0428, "step": 18220 }, { "grad_norm": 0.40139907598495483, "learning_rate": 9.159757868684727e-05, "loss": 0.0385, "step": 18230 }, { "grad_norm": 0.4627731740474701, "learning_rate": 9.15861073228243e-05, "loss": 0.0422, "step": 18240 }, { "grad_norm": 0.46418583393096924, "learning_rate": 9.157462885287296e-05, "loss": 0.0497, "step": 18250 }, { "grad_norm": 0.4307645857334137, "learning_rate": 9.156314327895461e-05, "loss": 0.0473, "step": 18260 }, { "grad_norm": 0.43687504529953003, "learning_rate": 9.155165060303185e-05, "loss": 0.043, "step": 18270 }, { "grad_norm": 0.44441843032836914, "learning_rate": 9.154015082706841e-05, "loss": 0.0416, "step": 18280 }, { "grad_norm": 0.34860706329345703, "learning_rate": 9.152864395302936e-05, "loss": 0.0419, "step": 18290 }, { "grad_norm": 0.4411437213420868, "learning_rate": 9.151712998288085e-05, "loss": 0.0374, "step": 18300 }, { "grad_norm": 0.5299049019813538, "learning_rate": 9.150560891859031e-05, "loss": 0.0436, "step": 18310 }, { "grad_norm": 0.4275750517845154, "learning_rate": 9.14940807621264e-05, "loss": 0.0406, "step": 18320 }, { "grad_norm": 0.45163628458976746, "learning_rate": 9.148254551545894e-05, "loss": 0.042, "step": 18330 }, { "grad_norm": 0.45889097452163696, "learning_rate": 9.147100318055901e-05, "loss": 0.0458, "step": 18340 }, { "grad_norm": 0.44534173607826233, "learning_rate": 9.145945375939888e-05, "loss": 0.0426, "step": 18350 }, { "grad_norm": 0.4452958405017853, "learning_rate": 9.144789725395203e-05, "loss": 0.0468, "step": 18360 }, { "grad_norm": 0.4560033977031708, "learning_rate": 9.14363336661931e-05, "loss": 0.0404, "step": 18370 }, { "grad_norm": 0.5114017128944397, "learning_rate": 9.142476299809806e-05, "loss": 0.0482, "step": 18380 }, { "grad_norm": 0.4213015139102936, "learning_rate": 9.1413185251644e-05, "loss": 0.0466, "step": 18390 }, { "grad_norm": 0.48275333642959595, "learning_rate": 9.140160042880923e-05, "loss": 0.0552, "step": 18400 }, { "grad_norm": 0.5088806748390198, "learning_rate": 9.139000853157327e-05, "loss": 0.0528, "step": 18410 }, { "grad_norm": 0.5038992762565613, "learning_rate": 9.137840956191688e-05, "loss": 0.0475, "step": 18420 }, { "grad_norm": 0.3717940151691437, "learning_rate": 9.136680352182199e-05, "loss": 0.0457, "step": 18430 }, { "grad_norm": 0.32466503977775574, "learning_rate": 9.135519041327177e-05, "loss": 0.0436, "step": 18440 }, { "grad_norm": 0.4122834801673889, "learning_rate": 9.134357023825058e-05, "loss": 0.0527, "step": 18450 }, { "grad_norm": 0.44335225224494934, "learning_rate": 9.133194299874398e-05, "loss": 0.0439, "step": 18460 }, { "grad_norm": 0.4803277552127838, "learning_rate": 9.132030869673876e-05, "loss": 0.0454, "step": 18470 }, { "grad_norm": 0.4001636207103729, "learning_rate": 9.130866733422288e-05, "loss": 0.0509, "step": 18480 }, { "grad_norm": 0.4656299352645874, "learning_rate": 9.129701891318556e-05, "loss": 0.0404, "step": 18490 }, { "grad_norm": 0.29568085074424744, "learning_rate": 9.128536343561718e-05, "loss": 0.0542, "step": 18500 }, { "grad_norm": 0.40121522545814514, "learning_rate": 9.127370090350934e-05, "loss": 0.0416, "step": 18510 }, { "grad_norm": 0.41384080052375793, "learning_rate": 9.126203131885487e-05, "loss": 0.0455, "step": 18520 }, { "grad_norm": 0.3744862675666809, "learning_rate": 9.125035468364775e-05, "loss": 0.0391, "step": 18530 }, { "grad_norm": 0.30139651894569397, "learning_rate": 9.123867099988322e-05, "loss": 0.0319, "step": 18540 }, { "grad_norm": 0.4001157879829407, "learning_rate": 9.122698026955769e-05, "loss": 0.0409, "step": 18550 }, { "grad_norm": 0.29466572403907776, "learning_rate": 9.12152824946688e-05, "loss": 0.065, "step": 18560 }, { "grad_norm": 0.47540709376335144, "learning_rate": 9.120357767721538e-05, "loss": 0.0473, "step": 18570 }, { "grad_norm": 0.5518321394920349, "learning_rate": 9.119186581919745e-05, "loss": 0.0425, "step": 18580 }, { "grad_norm": 0.5083851218223572, "learning_rate": 9.118014692261624e-05, "loss": 0.0402, "step": 18590 }, { "grad_norm": 0.45446619391441345, "learning_rate": 9.116842098947422e-05, "loss": 0.0581, "step": 18600 }, { "grad_norm": 0.34242528676986694, "learning_rate": 9.115668802177499e-05, "loss": 0.0369, "step": 18610 }, { "grad_norm": 0.4326692819595337, "learning_rate": 9.114494802152342e-05, "loss": 0.0526, "step": 18620 }, { "grad_norm": 0.584215521812439, "learning_rate": 9.113320099072555e-05, "loss": 0.0675, "step": 18630 }, { "grad_norm": 0.4764653146266937, "learning_rate": 9.112144693138864e-05, "loss": 0.0451, "step": 18640 }, { "grad_norm": 0.5359700918197632, "learning_rate": 9.110968584552111e-05, "loss": 0.038, "step": 18650 }, { "grad_norm": 0.3121464252471924, "learning_rate": 9.109791773513264e-05, "loss": 0.0464, "step": 18660 }, { "grad_norm": 0.38102078437805176, "learning_rate": 9.108614260223403e-05, "loss": 0.0433, "step": 18670 }, { "grad_norm": 0.4866960644721985, "learning_rate": 9.107436044883738e-05, "loss": 0.0444, "step": 18680 }, { "grad_norm": 0.48740366101264954, "learning_rate": 9.10625712769559e-05, "loss": 0.0403, "step": 18690 }, { "grad_norm": 0.46937012672424316, "learning_rate": 9.105077508860406e-05, "loss": 0.0378, "step": 18700 }, { "grad_norm": 0.3978305160999298, "learning_rate": 9.103897188579751e-05, "loss": 0.0376, "step": 18710 }, { "grad_norm": 0.3446654677391052, "learning_rate": 9.102716167055308e-05, "loss": 0.0345, "step": 18720 }, { "grad_norm": 0.4917483925819397, "learning_rate": 9.10153444448888e-05, "loss": 0.065, "step": 18730 }, { "grad_norm": 0.44358399510383606, "learning_rate": 9.100352021082393e-05, "loss": 0.0403, "step": 18740 }, { "grad_norm": 0.460956871509552, "learning_rate": 9.099168897037891e-05, "loss": 0.0399, "step": 18750 }, { "grad_norm": 0.538805365562439, "learning_rate": 9.097985072557538e-05, "loss": 0.0398, "step": 18760 }, { "grad_norm": 0.3795205354690552, "learning_rate": 9.096800547843615e-05, "loss": 0.0646, "step": 18770 }, { "grad_norm": 0.46232834458351135, "learning_rate": 9.095615323098526e-05, "loss": 0.0497, "step": 18780 }, { "grad_norm": 0.45091596245765686, "learning_rate": 9.094429398524795e-05, "loss": 0.0385, "step": 18790 }, { "grad_norm": 0.37264421582221985, "learning_rate": 9.093242774325061e-05, "loss": 0.0408, "step": 18800 }, { "grad_norm": 0.29406407475471497, "learning_rate": 9.092055450702088e-05, "loss": 0.0375, "step": 18810 }, { "grad_norm": 0.36848682165145874, "learning_rate": 9.090867427858756e-05, "loss": 0.0371, "step": 18820 }, { "grad_norm": 0.48916149139404297, "learning_rate": 9.089678705998066e-05, "loss": 0.0366, "step": 18830 }, { "grad_norm": 0.45073026418685913, "learning_rate": 9.088489285323139e-05, "loss": 0.041, "step": 18840 }, { "grad_norm": 0.397145539522171, "learning_rate": 9.087299166037212e-05, "loss": 0.0431, "step": 18850 }, { "grad_norm": 0.4957033693790436, "learning_rate": 9.086108348343647e-05, "loss": 0.0479, "step": 18860 }, { "grad_norm": 0.43937423825263977, "learning_rate": 9.08491683244592e-05, "loss": 0.0461, "step": 18870 }, { "grad_norm": 0.2528272867202759, "learning_rate": 9.08372461854763e-05, "loss": 0.0383, "step": 18880 }, { "grad_norm": 0.5537160038948059, "learning_rate": 9.082531706852492e-05, "loss": 0.043, "step": 18890 }, { "grad_norm": 0.3731114864349365, "learning_rate": 9.081338097564342e-05, "loss": 0.0497, "step": 18900 }, { "grad_norm": 0.4804636538028717, "learning_rate": 9.080143790887137e-05, "loss": 0.0418, "step": 18910 }, { "grad_norm": 0.4589819312095642, "learning_rate": 9.07894878702495e-05, "loss": 0.0347, "step": 18920 }, { "grad_norm": 0.3440675139427185, "learning_rate": 9.077753086181974e-05, "loss": 0.0482, "step": 18930 }, { "grad_norm": 0.3497462272644043, "learning_rate": 9.076556688562524e-05, "loss": 0.0318, "step": 18940 }, { "grad_norm": 0.7157433032989502, "learning_rate": 9.075359594371029e-05, "loss": 0.0324, "step": 18950 }, { "grad_norm": 0.39022111892700195, "learning_rate": 9.07416180381204e-05, "loss": 0.0344, "step": 18960 }, { "grad_norm": 0.4714924395084381, "learning_rate": 9.072963317090228e-05, "loss": 0.0412, "step": 18970 }, { "grad_norm": 0.48260170221328735, "learning_rate": 9.071764134410382e-05, "loss": 0.0393, "step": 18980 }, { "grad_norm": 0.5057623386383057, "learning_rate": 9.070564255977407e-05, "loss": 0.0343, "step": 18990 }, { "grad_norm": 0.4820176959037781, "learning_rate": 9.06936368199633e-05, "loss": 0.0537, "step": 19000 }, { "grad_norm": 0.408683180809021, "learning_rate": 9.0681624126723e-05, "loss": 0.0513, "step": 19010 }, { "grad_norm": 0.44960838556289673, "learning_rate": 9.066960448210576e-05, "loss": 0.0416, "step": 19020 }, { "grad_norm": 0.6011829376220703, "learning_rate": 9.065757788816543e-05, "loss": 0.0389, "step": 19030 }, { "grad_norm": 0.5621656775474548, "learning_rate": 9.064554434695705e-05, "loss": 0.0556, "step": 19040 }, { "grad_norm": 0.4137372672557831, "learning_rate": 9.063350386053677e-05, "loss": 0.048, "step": 19050 }, { "grad_norm": 0.45939335227012634, "learning_rate": 9.062145643096202e-05, "loss": 0.0397, "step": 19060 }, { "grad_norm": 0.39071887731552124, "learning_rate": 9.060940206029136e-05, "loss": 0.0447, "step": 19070 }, { "grad_norm": 0.4159339964389801, "learning_rate": 9.059734075058457e-05, "loss": 0.0451, "step": 19080 }, { "grad_norm": 0.9923442602157593, "learning_rate": 9.058527250390257e-05, "loss": 0.0558, "step": 19090 }, { "grad_norm": 0.3637405037879944, "learning_rate": 9.057319732230752e-05, "loss": 0.0633, "step": 19100 }, { "grad_norm": 0.35849905014038086, "learning_rate": 9.056111520786273e-05, "loss": 0.0461, "step": 19110 }, { "grad_norm": 0.3531343936920166, "learning_rate": 9.054902616263268e-05, "loss": 0.0424, "step": 19120 }, { "grad_norm": 0.6661380529403687, "learning_rate": 9.05369301886831e-05, "loss": 0.0586, "step": 19130 }, { "grad_norm": 0.5517428517341614, "learning_rate": 9.052482728808083e-05, "loss": 0.0466, "step": 19140 }, { "grad_norm": 0.34565067291259766, "learning_rate": 9.051271746289391e-05, "loss": 0.0275, "step": 19150 }, { "grad_norm": 0.33706533908843994, "learning_rate": 9.050060071519162e-05, "loss": 0.044, "step": 19160 }, { "grad_norm": 0.3360237777233124, "learning_rate": 9.048847704704437e-05, "loss": 0.035, "step": 19170 }, { "grad_norm": 0.3989441692829132, "learning_rate": 9.047634646052376e-05, "loss": 0.0519, "step": 19180 }, { "grad_norm": 0.5166197419166565, "learning_rate": 9.046420895770256e-05, "loss": 0.0556, "step": 19190 }, { "grad_norm": 0.2968968451023102, "learning_rate": 9.045206454065473e-05, "loss": 0.0355, "step": 19200 }, { "grad_norm": 0.3949500620365143, "learning_rate": 9.043991321145546e-05, "loss": 0.0346, "step": 19210 }, { "grad_norm": 0.46481287479400635, "learning_rate": 9.042775497218105e-05, "loss": 0.0389, "step": 19220 }, { "grad_norm": 0.44702577590942383, "learning_rate": 9.041558982490901e-05, "loss": 0.0446, "step": 19230 }, { "grad_norm": 0.4793374836444855, "learning_rate": 9.040341777171805e-05, "loss": 0.0445, "step": 19240 }, { "grad_norm": 0.44315263628959656, "learning_rate": 9.039123881468802e-05, "loss": 0.0376, "step": 19250 }, { "grad_norm": 0.4272855520248413, "learning_rate": 9.037905295589998e-05, "loss": 0.0441, "step": 19260 }, { "grad_norm": 0.434285432100296, "learning_rate": 9.036686019743617e-05, "loss": 0.0364, "step": 19270 }, { "grad_norm": 0.3852582275867462, "learning_rate": 9.035466054137997e-05, "loss": 0.039, "step": 19280 }, { "grad_norm": 0.34208568930625916, "learning_rate": 9.0342453989816e-05, "loss": 0.0471, "step": 19290 }, { "grad_norm": 0.30075255036354065, "learning_rate": 9.033024054483e-05, "loss": 0.0408, "step": 19300 }, { "grad_norm": 0.4055686891078949, "learning_rate": 9.031802020850894e-05, "loss": 0.0408, "step": 19310 }, { "grad_norm": 0.4489266872406006, "learning_rate": 9.030579298294092e-05, "loss": 0.0491, "step": 19320 }, { "grad_norm": 0.37906384468078613, "learning_rate": 9.029355887021524e-05, "loss": 0.0503, "step": 19330 }, { "grad_norm": 0.5484864115715027, "learning_rate": 9.028131787242238e-05, "loss": 0.0614, "step": 19340 }, { "grad_norm": 0.33475232124328613, "learning_rate": 9.026906999165399e-05, "loss": 0.0466, "step": 19350 }, { "grad_norm": 0.33109015226364136, "learning_rate": 9.025681523000291e-05, "loss": 0.0341, "step": 19360 }, { "grad_norm": 0.48584678769111633, "learning_rate": 9.024455358956315e-05, "loss": 0.0382, "step": 19370 }, { "grad_norm": 0.38835158944129944, "learning_rate": 9.023228507242984e-05, "loss": 0.0434, "step": 19380 }, { "grad_norm": 0.3103514611721039, "learning_rate": 9.022000968069937e-05, "loss": 0.0372, "step": 19390 }, { "grad_norm": 0.4616529941558838, "learning_rate": 9.020772741646928e-05, "loss": 0.037, "step": 19400 }, { "grad_norm": 0.4397343099117279, "learning_rate": 9.019543828183826e-05, "loss": 0.0337, "step": 19410 }, { "grad_norm": 0.3682789206504822, "learning_rate": 9.018314227890616e-05, "loss": 0.0381, "step": 19420 }, { "grad_norm": 0.47011423110961914, "learning_rate": 9.017083940977408e-05, "loss": 0.0516, "step": 19430 }, { "grad_norm": 0.3492841124534607, "learning_rate": 9.015852967654422e-05, "loss": 0.0391, "step": 19440 }, { "grad_norm": 0.3869985044002533, "learning_rate": 9.014621308131996e-05, "loss": 0.0304, "step": 19450 }, { "grad_norm": 0.4632968008518219, "learning_rate": 9.01338896262059e-05, "loss": 0.0392, "step": 19460 }, { "grad_norm": 0.3708793520927429, "learning_rate": 9.012155931330777e-05, "loss": 0.0465, "step": 19470 }, { "grad_norm": 0.39678022265434265, "learning_rate": 9.010922214473246e-05, "loss": 0.0329, "step": 19480 }, { "grad_norm": 0.3647286295890808, "learning_rate": 9.009687812258808e-05, "loss": 0.0525, "step": 19490 }, { "grad_norm": 0.3819924294948578, "learning_rate": 9.00845272489839e-05, "loss": 0.0444, "step": 19500 }, { "grad_norm": 0.34434446692466736, "learning_rate": 9.007216952603031e-05, "loss": 0.0522, "step": 19510 }, { "grad_norm": 0.27336663007736206, "learning_rate": 9.005980495583894e-05, "loss": 0.037, "step": 19520 }, { "grad_norm": 0.4554177224636078, "learning_rate": 9.004743354052252e-05, "loss": 0.0422, "step": 19530 }, { "grad_norm": 0.4472986161708832, "learning_rate": 9.003505528219503e-05, "loss": 0.0354, "step": 19540 }, { "grad_norm": 0.3904343843460083, "learning_rate": 9.002267018297154e-05, "loss": 0.042, "step": 19550 }, { "grad_norm": 0.3786570429801941, "learning_rate": 9.001027824496834e-05, "loss": 0.0421, "step": 19560 }, { "grad_norm": 0.3326028883457184, "learning_rate": 8.999787947030287e-05, "loss": 0.0393, "step": 19570 }, { "grad_norm": 0.4431619644165039, "learning_rate": 8.998547386109376e-05, "loss": 0.038, "step": 19580 }, { "grad_norm": 0.44668474793434143, "learning_rate": 8.997306141946073e-05, "loss": 0.0402, "step": 19590 }, { "grad_norm": 0.544761598110199, "learning_rate": 8.996064214752481e-05, "loss": 0.0426, "step": 19600 }, { "grad_norm": 0.4294208288192749, "learning_rate": 8.994821604740806e-05, "loss": 0.0424, "step": 19610 }, { "grad_norm": 0.4457815885543823, "learning_rate": 8.993578312123377e-05, "loss": 0.0392, "step": 19620 }, { "grad_norm": 0.32609137892723083, "learning_rate": 8.992334337112639e-05, "loss": 0.045, "step": 19630 }, { "grad_norm": 0.3731541037559509, "learning_rate": 8.991089679921154e-05, "loss": 0.0306, "step": 19640 }, { "grad_norm": 0.4775170385837555, "learning_rate": 8.989844340761599e-05, "loss": 0.0385, "step": 19650 }, { "grad_norm": 0.4690499007701874, "learning_rate": 8.988598319846768e-05, "loss": 0.0372, "step": 19660 }, { "grad_norm": 0.3095918297767639, "learning_rate": 8.987351617389574e-05, "loss": 0.0344, "step": 19670 }, { "grad_norm": 0.34632572531700134, "learning_rate": 8.98610423360304e-05, "loss": 0.0343, "step": 19680 }, { "grad_norm": 0.2716900706291199, "learning_rate": 8.984856168700317e-05, "loss": 0.0455, "step": 19690 }, { "grad_norm": 0.3952119052410126, "learning_rate": 8.983607422894658e-05, "loss": 0.0402, "step": 19700 }, { "grad_norm": 0.3330661356449127, "learning_rate": 8.982357996399442e-05, "loss": 0.0454, "step": 19710 }, { "grad_norm": 0.4513315260410309, "learning_rate": 8.981107889428164e-05, "loss": 0.0465, "step": 19720 }, { "grad_norm": 0.3582213222980499, "learning_rate": 8.979857102194428e-05, "loss": 0.0344, "step": 19730 }, { "grad_norm": 0.7711641192436218, "learning_rate": 8.978605634911968e-05, "loss": 0.0372, "step": 19740 }, { "grad_norm": 0.43205493688583374, "learning_rate": 8.977353487794616e-05, "loss": 0.0312, "step": 19750 }, { "grad_norm": 0.571789026260376, "learning_rate": 8.976100661056334e-05, "loss": 0.0501, "step": 19760 }, { "grad_norm": 0.4091322422027588, "learning_rate": 8.974847154911197e-05, "loss": 0.0352, "step": 19770 }, { "grad_norm": 0.4787120223045349, "learning_rate": 8.973592969573393e-05, "loss": 0.0482, "step": 19780 }, { "grad_norm": 0.48349860310554504, "learning_rate": 8.972338105257228e-05, "loss": 0.0502, "step": 19790 }, { "grad_norm": 0.3340280055999756, "learning_rate": 8.971082562177125e-05, "loss": 0.0399, "step": 19800 }, { "grad_norm": 0.5076205134391785, "learning_rate": 8.96982634054762e-05, "loss": 0.0386, "step": 19810 }, { "grad_norm": 0.3469471335411072, "learning_rate": 8.96856944058337e-05, "loss": 0.035, "step": 19820 }, { "grad_norm": 0.29613441228866577, "learning_rate": 8.967311862499144e-05, "loss": 0.0254, "step": 19830 }, { "grad_norm": 0.2663448750972748, "learning_rate": 8.966053606509825e-05, "loss": 0.0355, "step": 19840 }, { "grad_norm": 0.4630231261253357, "learning_rate": 8.964794672830417e-05, "loss": 0.0453, "step": 19850 }, { "grad_norm": 0.46139833331108093, "learning_rate": 8.963535061676038e-05, "loss": 0.0538, "step": 19860 }, { "grad_norm": 0.41313427686691284, "learning_rate": 8.962274773261918e-05, "loss": 0.0418, "step": 19870 }, { "grad_norm": 0.5651108026504517, "learning_rate": 8.961013807803409e-05, "loss": 0.0493, "step": 19880 }, { "grad_norm": 0.5444294810295105, "learning_rate": 8.959752165515973e-05, "loss": 0.0461, "step": 19890 }, { "grad_norm": 0.5681939721107483, "learning_rate": 8.958489846615193e-05, "loss": 0.0589, "step": 19900 }, { "grad_norm": 0.5283496379852295, "learning_rate": 8.957226851316762e-05, "loss": 0.0392, "step": 19910 }, { "grad_norm": 0.3809511661529541, "learning_rate": 8.955963179836493e-05, "loss": 0.045, "step": 19920 }, { "grad_norm": 0.29009127616882324, "learning_rate": 8.954698832390312e-05, "loss": 0.0388, "step": 19930 }, { "grad_norm": 0.4935983717441559, "learning_rate": 8.953433809194263e-05, "loss": 0.0374, "step": 19940 }, { "grad_norm": 0.34068864583969116, "learning_rate": 8.9521681104645e-05, "loss": 0.0391, "step": 19950 }, { "grad_norm": 0.550993800163269, "learning_rate": 8.9509017364173e-05, "loss": 0.0473, "step": 19960 }, { "grad_norm": 0.5219205021858215, "learning_rate": 8.949634687269052e-05, "loss": 0.04, "step": 19970 }, { "grad_norm": 0.477500319480896, "learning_rate": 8.948366963236259e-05, "loss": 0.053, "step": 19980 }, { "grad_norm": 0.3571743667125702, "learning_rate": 8.947098564535538e-05, "loss": 0.0406, "step": 19990 }, { "grad_norm": 0.42827942967414856, "learning_rate": 8.945829491383627e-05, "loss": 0.0433, "step": 20000 }, { "grad_norm": 0.40836256742477417, "learning_rate": 8.944559743997374e-05, "loss": 0.0408, "step": 20010 }, { "grad_norm": 0.25857895612716675, "learning_rate": 8.943289322593746e-05, "loss": 0.0393, "step": 20020 }, { "grad_norm": 0.420767605304718, "learning_rate": 8.942018227389821e-05, "loss": 0.0387, "step": 20030 }, { "grad_norm": 0.4380788803100586, "learning_rate": 8.940746458602795e-05, "loss": 0.0582, "step": 20040 }, { "grad_norm": 0.5427178144454956, "learning_rate": 8.939474016449979e-05, "loss": 0.0529, "step": 20050 }, { "grad_norm": 0.36180126667022705, "learning_rate": 8.938200901148799e-05, "loss": 0.039, "step": 20060 }, { "grad_norm": 0.5892043113708496, "learning_rate": 8.936927112916795e-05, "loss": 0.0485, "step": 20070 }, { "grad_norm": 0.49372169375419617, "learning_rate": 8.935652651971622e-05, "loss": 0.0359, "step": 20080 }, { "grad_norm": 0.5376598238945007, "learning_rate": 8.934377518531052e-05, "loss": 0.0386, "step": 20090 }, { "grad_norm": 0.5314151644706726, "learning_rate": 8.933101712812967e-05, "loss": 0.0409, "step": 20100 }, { "grad_norm": 0.43938761949539185, "learning_rate": 8.931825235035374e-05, "loss": 0.0442, "step": 20110 }, { "grad_norm": 0.31649547815322876, "learning_rate": 8.930548085416382e-05, "loss": 0.0391, "step": 20120 }, { "grad_norm": 0.8522511720657349, "learning_rate": 8.92927026417422e-05, "loss": 0.044, "step": 20130 }, { "grad_norm": 0.43250325322151184, "learning_rate": 8.92799177152724e-05, "loss": 0.0381, "step": 20140 }, { "grad_norm": 0.3554348647594452, "learning_rate": 8.926712607693895e-05, "loss": 0.04, "step": 20150 }, { "grad_norm": 0.35134410858154297, "learning_rate": 8.925432772892762e-05, "loss": 0.0475, "step": 20160 }, { "grad_norm": 0.598685085773468, "learning_rate": 8.924152267342529e-05, "loss": 0.0516, "step": 20170 }, { "grad_norm": 0.43225663900375366, "learning_rate": 8.922871091261998e-05, "loss": 0.0466, "step": 20180 }, { "grad_norm": 0.30345505475997925, "learning_rate": 8.92158924487009e-05, "loss": 0.038, "step": 20190 }, { "grad_norm": 0.35568663477897644, "learning_rate": 8.920306728385834e-05, "loss": 0.0335, "step": 20200 }, { "grad_norm": 0.6365461349487305, "learning_rate": 8.919023542028379e-05, "loss": 0.0344, "step": 20210 }, { "grad_norm": 0.3255167007446289, "learning_rate": 8.917739686016988e-05, "loss": 0.0387, "step": 20220 }, { "grad_norm": 0.4588681757450104, "learning_rate": 8.916455160571033e-05, "loss": 0.0519, "step": 20230 }, { "grad_norm": 0.40093234181404114, "learning_rate": 8.915169965910008e-05, "loss": 0.0322, "step": 20240 }, { "grad_norm": 0.43356212973594666, "learning_rate": 8.913884102253514e-05, "loss": 0.0352, "step": 20250 }, { "grad_norm": 0.3494287133216858, "learning_rate": 8.912597569821273e-05, "loss": 0.0316, "step": 20260 }, { "grad_norm": 0.3515125513076782, "learning_rate": 8.911310368833118e-05, "loss": 0.0384, "step": 20270 }, { "grad_norm": 0.43664366006851196, "learning_rate": 8.910022499508994e-05, "loss": 0.0313, "step": 20280 }, { "grad_norm": 0.33040115237236023, "learning_rate": 8.908733962068965e-05, "loss": 0.0409, "step": 20290 }, { "grad_norm": 0.3062252104282379, "learning_rate": 8.907444756733207e-05, "loss": 0.032, "step": 20300 }, { "grad_norm": 0.45270127058029175, "learning_rate": 8.906154883722006e-05, "loss": 0.0418, "step": 20310 }, { "grad_norm": 0.3445010483264923, "learning_rate": 8.904864343255773e-05, "loss": 0.037, "step": 20320 }, { "grad_norm": 0.4426952004432678, "learning_rate": 8.90357313555502e-05, "loss": 0.04, "step": 20330 }, { "grad_norm": 0.3251490592956543, "learning_rate": 8.90228126084038e-05, "loss": 0.0432, "step": 20340 }, { "grad_norm": 0.4103695750236511, "learning_rate": 8.900988719332601e-05, "loss": 0.0509, "step": 20350 }, { "grad_norm": 0.25285083055496216, "learning_rate": 8.899695511252542e-05, "loss": 0.0482, "step": 20360 }, { "grad_norm": 0.390048086643219, "learning_rate": 8.898401636821176e-05, "loss": 0.0508, "step": 20370 }, { "grad_norm": 0.5258376002311707, "learning_rate": 8.897107096259593e-05, "loss": 0.0547, "step": 20380 }, { "grad_norm": 0.40442365407943726, "learning_rate": 8.895811889788994e-05, "loss": 0.0506, "step": 20390 }, { "grad_norm": 0.4587138295173645, "learning_rate": 8.894516017630692e-05, "loss": 0.0472, "step": 20400 }, { "grad_norm": 0.4314010739326477, "learning_rate": 8.893219480006118e-05, "loss": 0.0441, "step": 20410 }, { "grad_norm": 0.46490010619163513, "learning_rate": 8.891922277136817e-05, "loss": 0.0434, "step": 20420 }, { "grad_norm": 0.29778921604156494, "learning_rate": 8.890624409244441e-05, "loss": 0.0359, "step": 20430 }, { "grad_norm": 0.34891176223754883, "learning_rate": 8.889325876550763e-05, "loss": 0.0432, "step": 20440 }, { "grad_norm": 0.29220277070999146, "learning_rate": 8.888026679277666e-05, "loss": 0.0361, "step": 20450 }, { "grad_norm": 0.28971824049949646, "learning_rate": 8.886726817647147e-05, "loss": 0.0597, "step": 20460 }, { "grad_norm": 0.268453449010849, "learning_rate": 8.885426291881319e-05, "loss": 0.0519, "step": 20470 }, { "grad_norm": 0.3877265155315399, "learning_rate": 8.884125102202401e-05, "loss": 0.041, "step": 20480 }, { "grad_norm": 0.3245287537574768, "learning_rate": 8.882823248832736e-05, "loss": 0.0277, "step": 20490 }, { "grad_norm": 0.4151252210140228, "learning_rate": 8.881520731994772e-05, "loss": 0.0365, "step": 20500 }, { "grad_norm": 0.34146466851234436, "learning_rate": 8.880217551911077e-05, "loss": 0.0371, "step": 20510 }, { "grad_norm": 0.39663371443748474, "learning_rate": 8.878913708804323e-05, "loss": 0.0321, "step": 20520 }, { "grad_norm": 0.3155779242515564, "learning_rate": 8.877609202897308e-05, "loss": 0.0453, "step": 20530 }, { "grad_norm": 0.3723747432231903, "learning_rate": 8.876304034412933e-05, "loss": 0.0315, "step": 20540 }, { "grad_norm": 0.46399056911468506, "learning_rate": 8.874998203574214e-05, "loss": 0.0475, "step": 20550 }, { "grad_norm": 0.4299602806568146, "learning_rate": 8.873691710604284e-05, "loss": 0.0386, "step": 20560 }, { "grad_norm": 0.4320726990699768, "learning_rate": 8.872384555726387e-05, "loss": 0.0389, "step": 20570 }, { "grad_norm": 0.371141642332077, "learning_rate": 8.871076739163878e-05, "loss": 0.0396, "step": 20580 }, { "grad_norm": 0.3497561514377594, "learning_rate": 8.86976826114023e-05, "loss": 0.0398, "step": 20590 }, { "grad_norm": 0.9213319420814514, "learning_rate": 8.868459121879023e-05, "loss": 0.0411, "step": 20600 }, { "grad_norm": 0.9845086932182312, "learning_rate": 8.867149321603956e-05, "loss": 0.0503, "step": 20610 }, { "grad_norm": 0.3424173593521118, "learning_rate": 8.865838860538835e-05, "loss": 0.0405, "step": 20620 }, { "grad_norm": 0.45606717467308044, "learning_rate": 8.864527738907585e-05, "loss": 0.0412, "step": 20630 }, { "grad_norm": 0.48272302746772766, "learning_rate": 8.863215956934239e-05, "loss": 0.051, "step": 20640 }, { "grad_norm": 0.35900917649269104, "learning_rate": 8.861903514842947e-05, "loss": 0.0336, "step": 20650 }, { "grad_norm": 0.3166324496269226, "learning_rate": 8.860590412857966e-05, "loss": 0.0349, "step": 20660 }, { "grad_norm": 0.40791720151901245, "learning_rate": 8.85927665120367e-05, "loss": 0.0379, "step": 20670 }, { "grad_norm": 0.4834573566913605, "learning_rate": 8.857962230104546e-05, "loss": 0.0369, "step": 20680 }, { "grad_norm": 0.4768321216106415, "learning_rate": 8.856647149785193e-05, "loss": 0.0404, "step": 20690 }, { "grad_norm": 0.32811227440834045, "learning_rate": 8.855331410470322e-05, "loss": 0.0434, "step": 20700 }, { "grad_norm": 0.4454309642314911, "learning_rate": 8.854015012384756e-05, "loss": 0.0368, "step": 20710 }, { "grad_norm": 0.566069483757019, "learning_rate": 8.852697955753433e-05, "loss": 0.0529, "step": 20720 }, { "grad_norm": 0.30763354897499084, "learning_rate": 8.851380240801399e-05, "loss": 0.0457, "step": 20730 }, { "grad_norm": 0.3431563973426819, "learning_rate": 8.850061867753818e-05, "loss": 0.0298, "step": 20740 }, { "grad_norm": 0.5738632082939148, "learning_rate": 8.848742836835963e-05, "loss": 0.0536, "step": 20750 }, { "grad_norm": 0.3515564203262329, "learning_rate": 8.847423148273221e-05, "loss": 0.0369, "step": 20760 }, { "grad_norm": 0.3592425286769867, "learning_rate": 8.846102802291092e-05, "loss": 0.0527, "step": 20770 }, { "grad_norm": 0.5152408480644226, "learning_rate": 8.844781799115183e-05, "loss": 0.0445, "step": 20780 }, { "grad_norm": 0.5194791555404663, "learning_rate": 8.84346013897122e-05, "loss": 0.0361, "step": 20790 }, { "grad_norm": 0.45124688744544983, "learning_rate": 8.842137822085038e-05, "loss": 0.0457, "step": 20800 }, { "grad_norm": 0.4105449616909027, "learning_rate": 8.840814848682585e-05, "loss": 0.0356, "step": 20810 }, { "grad_norm": 0.36673590540885925, "learning_rate": 8.83949121898992e-05, "loss": 0.0499, "step": 20820 }, { "grad_norm": 0.3949495255947113, "learning_rate": 8.838166933233217e-05, "loss": 0.0322, "step": 20830 }, { "grad_norm": 0.4020228385925293, "learning_rate": 8.83684199163876e-05, "loss": 0.0481, "step": 20840 }, { "grad_norm": 0.3398834764957428, "learning_rate": 8.835516394432943e-05, "loss": 0.0474, "step": 20850 }, { "grad_norm": 0.3776678740978241, "learning_rate": 8.834190141842276e-05, "loss": 0.038, "step": 20860 }, { "grad_norm": 0.47281208634376526, "learning_rate": 8.83286323409338e-05, "loss": 0.0339, "step": 20870 }, { "grad_norm": 0.3811306655406952, "learning_rate": 8.831535671412986e-05, "loss": 0.0416, "step": 20880 }, { "grad_norm": 0.36990079283714294, "learning_rate": 8.830207454027938e-05, "loss": 0.0403, "step": 20890 }, { "grad_norm": 0.3403024673461914, "learning_rate": 8.828878582165192e-05, "loss": 0.0396, "step": 20900 }, { "grad_norm": 0.3851338028907776, "learning_rate": 8.827549056051818e-05, "loss": 0.0391, "step": 20910 }, { "grad_norm": 0.4147108197212219, "learning_rate": 8.826218875914993e-05, "loss": 0.0369, "step": 20920 }, { "grad_norm": 0.3595805764198303, "learning_rate": 8.82488804198201e-05, "loss": 0.0339, "step": 20930 }, { "grad_norm": 0.29288819432258606, "learning_rate": 8.82355655448027e-05, "loss": 0.04, "step": 20940 }, { "grad_norm": 0.3881135582923889, "learning_rate": 8.822224413637293e-05, "loss": 0.044, "step": 20950 }, { "grad_norm": 0.40069639682769775, "learning_rate": 8.820891619680697e-05, "loss": 0.0302, "step": 20960 }, { "grad_norm": 0.4368145167827606, "learning_rate": 8.819558172838227e-05, "loss": 0.0371, "step": 20970 }, { "grad_norm": 0.3536880612373352, "learning_rate": 8.818224073337731e-05, "loss": 0.03, "step": 20980 }, { "grad_norm": 0.3467693030834198, "learning_rate": 8.816889321407169e-05, "loss": 0.0368, "step": 20990 }, { "grad_norm": 0.5904335975646973, "learning_rate": 8.815553917274615e-05, "loss": 0.0453, "step": 21000 }, { "grad_norm": 0.4048464000225067, "learning_rate": 8.81421786116825e-05, "loss": 0.0467, "step": 21010 }, { "grad_norm": 0.3964890241622925, "learning_rate": 8.812881153316373e-05, "loss": 0.0348, "step": 21020 }, { "grad_norm": 0.264631062746048, "learning_rate": 8.81154379394739e-05, "loss": 0.0434, "step": 21030 }, { "grad_norm": 0.4491651952266693, "learning_rate": 8.810205783289818e-05, "loss": 0.0414, "step": 21040 }, { "grad_norm": 0.48877233266830444, "learning_rate": 8.808867121572286e-05, "loss": 0.0442, "step": 21050 }, { "grad_norm": 0.3461568355560303, "learning_rate": 8.807527809023537e-05, "loss": 0.0437, "step": 21060 }, { "grad_norm": 0.4087429344654083, "learning_rate": 8.80618784587242e-05, "loss": 0.0432, "step": 21070 }, { "grad_norm": 0.3078637421131134, "learning_rate": 8.804847232347902e-05, "loss": 0.0392, "step": 21080 }, { "grad_norm": 0.8412247896194458, "learning_rate": 8.803505968679054e-05, "loss": 0.0515, "step": 21090 }, { "grad_norm": 0.4602518081665039, "learning_rate": 8.802164055095061e-05, "loss": 0.0473, "step": 21100 }, { "grad_norm": 0.4773752987384796, "learning_rate": 8.80082149182522e-05, "loss": 0.0366, "step": 21110 }, { "grad_norm": 0.3444719910621643, "learning_rate": 8.79947827909894e-05, "loss": 0.036, "step": 21120 }, { "grad_norm": 0.41921761631965637, "learning_rate": 8.798134417145738e-05, "loss": 0.0458, "step": 21130 }, { "grad_norm": 0.38568466901779175, "learning_rate": 8.796789906195243e-05, "loss": 0.0503, "step": 21140 }, { "grad_norm": 0.3987686336040497, "learning_rate": 8.795444746477195e-05, "loss": 0.0402, "step": 21150 }, { "grad_norm": 0.45787638425827026, "learning_rate": 8.794098938221446e-05, "loss": 0.0378, "step": 21160 }, { "grad_norm": 0.5902227163314819, "learning_rate": 8.792752481657957e-05, "loss": 0.0377, "step": 21170 }, { "grad_norm": 0.3611835241317749, "learning_rate": 8.791405377016802e-05, "loss": 0.0296, "step": 21180 }, { "grad_norm": 0.5034393668174744, "learning_rate": 8.790057624528163e-05, "loss": 0.0404, "step": 21190 }, { "grad_norm": 0.33950275182724, "learning_rate": 8.788709224422333e-05, "loss": 0.0355, "step": 21200 }, { "grad_norm": 0.33273792266845703, "learning_rate": 8.787360176929717e-05, "loss": 0.0373, "step": 21210 }, { "grad_norm": 0.34659966826438904, "learning_rate": 8.786010482280834e-05, "loss": 0.0353, "step": 21220 }, { "grad_norm": 0.37812408804893494, "learning_rate": 8.784660140706306e-05, "loss": 0.0414, "step": 21230 }, { "grad_norm": 0.31997087597846985, "learning_rate": 8.783309152436872e-05, "loss": 0.0288, "step": 21240 }, { "grad_norm": 0.3719102144241333, "learning_rate": 8.781957517703375e-05, "loss": 0.0413, "step": 21250 }, { "grad_norm": 0.5306218862533569, "learning_rate": 8.780605236736776e-05, "loss": 0.045, "step": 21260 }, { "grad_norm": 0.2971860468387604, "learning_rate": 8.779252309768142e-05, "loss": 0.0399, "step": 21270 }, { "grad_norm": 0.41474878787994385, "learning_rate": 8.777898737028652e-05, "loss": 0.0423, "step": 21280 }, { "grad_norm": 0.36719754338264465, "learning_rate": 8.776544518749591e-05, "loss": 0.0349, "step": 21290 }, { "grad_norm": 0.420894980430603, "learning_rate": 8.775189655162364e-05, "loss": 0.0345, "step": 21300 }, { "grad_norm": 0.4597829282283783, "learning_rate": 8.773834146498474e-05, "loss": 0.0376, "step": 21310 }, { "grad_norm": 0.46698933839797974, "learning_rate": 8.772477992989545e-05, "loss": 0.039, "step": 21320 }, { "grad_norm": 0.9046273231506348, "learning_rate": 8.771121194867304e-05, "loss": 0.0447, "step": 21330 }, { "grad_norm": 0.27142617106437683, "learning_rate": 8.769763752363589e-05, "loss": 0.0398, "step": 21340 }, { "grad_norm": 0.43920060992240906, "learning_rate": 8.768405665710352e-05, "loss": 0.0402, "step": 21350 }, { "grad_norm": 0.48394519090652466, "learning_rate": 8.767046935139655e-05, "loss": 0.0478, "step": 21360 }, { "grad_norm": 0.5694006085395813, "learning_rate": 8.765687560883666e-05, "loss": 0.0401, "step": 21370 }, { "grad_norm": 0.3818609118461609, "learning_rate": 8.764327543174664e-05, "loss": 0.036, "step": 21380 }, { "grad_norm": 0.5089694261550903, "learning_rate": 8.762966882245038e-05, "loss": 0.0393, "step": 21390 }, { "grad_norm": 0.335849404335022, "learning_rate": 8.761605578327291e-05, "loss": 0.0435, "step": 21400 }, { "grad_norm": 0.31112322211265564, "learning_rate": 8.76024363165403e-05, "loss": 0.047, "step": 21410 }, { "grad_norm": 0.3853467106819153, "learning_rate": 8.758881042457976e-05, "loss": 0.0385, "step": 21420 }, { "grad_norm": 0.4809703230857849, "learning_rate": 8.757517810971957e-05, "loss": 0.0356, "step": 21430 }, { "grad_norm": 0.5064233541488647, "learning_rate": 8.756153937428913e-05, "loss": 0.0479, "step": 21440 }, { "grad_norm": 0.3091396987438202, "learning_rate": 8.754789422061889e-05, "loss": 0.0438, "step": 21450 }, { "grad_norm": 0.3214589059352875, "learning_rate": 8.753424265104052e-05, "loss": 0.0387, "step": 21460 }, { "grad_norm": 0.4062098264694214, "learning_rate": 8.752058466788659e-05, "loss": 0.0453, "step": 21470 }, { "grad_norm": 0.451083242893219, "learning_rate": 8.750692027349097e-05, "loss": 0.0318, "step": 21480 }, { "grad_norm": 0.38080716133117676, "learning_rate": 8.749324947018847e-05, "loss": 0.0353, "step": 21490 }, { "grad_norm": 0.38527652621269226, "learning_rate": 8.747957226031507e-05, "loss": 0.0333, "step": 21500 }, { "grad_norm": 0.39216238260269165, "learning_rate": 8.746588864620787e-05, "loss": 0.0436, "step": 21510 }, { "grad_norm": 0.33016645908355713, "learning_rate": 8.745219863020498e-05, "loss": 0.0298, "step": 21520 }, { "grad_norm": 0.3376941680908203, "learning_rate": 8.743850221464564e-05, "loss": 0.0297, "step": 21530 }, { "grad_norm": 0.24741247296333313, "learning_rate": 8.742479940187026e-05, "loss": 0.0368, "step": 21540 }, { "grad_norm": 0.3466784954071045, "learning_rate": 8.74110901942202e-05, "loss": 0.0327, "step": 21550 }, { "grad_norm": 0.407565176486969, "learning_rate": 8.739737459403803e-05, "loss": 0.0373, "step": 21560 }, { "grad_norm": 0.25054073333740234, "learning_rate": 8.738365260366737e-05, "loss": 0.0379, "step": 21570 }, { "grad_norm": 0.36001354455947876, "learning_rate": 8.736992422545292e-05, "loss": 0.0274, "step": 21580 }, { "grad_norm": 0.43938177824020386, "learning_rate": 8.73561894617405e-05, "loss": 0.048, "step": 21590 }, { "grad_norm": 0.3721523582935333, "learning_rate": 8.734244831487697e-05, "loss": 0.0392, "step": 21600 }, { "grad_norm": 0.43656566739082336, "learning_rate": 8.732870078721035e-05, "loss": 0.045, "step": 21610 }, { "grad_norm": 0.3013240396976471, "learning_rate": 8.731494688108972e-05, "loss": 0.0387, "step": 21620 }, { "grad_norm": 0.6812883615493774, "learning_rate": 8.730118659886523e-05, "loss": 0.0394, "step": 21630 }, { "grad_norm": 0.40667563676834106, "learning_rate": 8.728741994288814e-05, "loss": 0.0388, "step": 21640 }, { "grad_norm": 0.3238948881626129, "learning_rate": 8.727364691551079e-05, "loss": 0.0371, "step": 21650 }, { "grad_norm": 0.34608644247055054, "learning_rate": 8.725986751908661e-05, "loss": 0.0444, "step": 21660 }, { "grad_norm": 0.45972514152526855, "learning_rate": 8.724608175597016e-05, "loss": 0.038, "step": 21670 }, { "grad_norm": 0.273425817489624, "learning_rate": 8.723228962851699e-05, "loss": 0.0342, "step": 21680 }, { "grad_norm": 0.41090014576911926, "learning_rate": 8.721849113908385e-05, "loss": 0.0379, "step": 21690 }, { "grad_norm": 0.5083465576171875, "learning_rate": 8.720468629002848e-05, "loss": 0.0376, "step": 21700 }, { "grad_norm": 0.4020792543888092, "learning_rate": 8.719087508370978e-05, "loss": 0.0347, "step": 21710 }, { "grad_norm": 0.5839852690696716, "learning_rate": 8.717705752248772e-05, "loss": 0.0501, "step": 21720 }, { "grad_norm": 0.5503195524215698, "learning_rate": 8.71632336087233e-05, "loss": 0.0342, "step": 21730 }, { "grad_norm": 0.4132448732852936, "learning_rate": 8.71494033447787e-05, "loss": 0.0381, "step": 21740 }, { "grad_norm": 0.30045443773269653, "learning_rate": 8.713556673301708e-05, "loss": 0.034, "step": 21750 }, { "grad_norm": 0.31365177035331726, "learning_rate": 8.712172377580278e-05, "loss": 0.0473, "step": 21760 }, { "grad_norm": 0.42240047454833984, "learning_rate": 8.710787447550114e-05, "loss": 0.0404, "step": 21770 }, { "grad_norm": 0.5067658424377441, "learning_rate": 8.70940188344787e-05, "loss": 0.061, "step": 21780 }, { "grad_norm": 0.39962247014045715, "learning_rate": 8.708015685510293e-05, "loss": 0.0395, "step": 21790 }, { "grad_norm": 0.32428795099258423, "learning_rate": 8.706628853974252e-05, "loss": 0.0454, "step": 21800 }, { "grad_norm": 0.4096425175666809, "learning_rate": 8.705241389076715e-05, "loss": 0.0331, "step": 21810 }, { "grad_norm": 0.29297932982444763, "learning_rate": 8.703853291054764e-05, "loss": 0.0324, "step": 21820 }, { "grad_norm": 0.4466114044189453, "learning_rate": 8.702464560145587e-05, "loss": 0.0417, "step": 21830 }, { "grad_norm": 0.2959767282009125, "learning_rate": 8.701075196586476e-05, "loss": 0.0353, "step": 21840 }, { "grad_norm": 0.3395126461982727, "learning_rate": 8.699685200614842e-05, "loss": 0.0335, "step": 21850 }, { "grad_norm": 0.3149096667766571, "learning_rate": 8.698294572468193e-05, "loss": 0.0411, "step": 21860 }, { "grad_norm": 0.27987173199653625, "learning_rate": 8.696903312384148e-05, "loss": 0.0349, "step": 21870 }, { "grad_norm": 0.3413812220096588, "learning_rate": 8.695511420600439e-05, "loss": 0.0276, "step": 21880 }, { "grad_norm": 0.3401176631450653, "learning_rate": 8.694118897354901e-05, "loss": 0.0298, "step": 21890 }, { "grad_norm": 0.44525811076164246, "learning_rate": 8.692725742885478e-05, "loss": 0.0353, "step": 21900 }, { "grad_norm": 0.5608786344528198, "learning_rate": 8.691331957430221e-05, "loss": 0.0394, "step": 21910 }, { "grad_norm": 0.3419318199157715, "learning_rate": 8.68993754122729e-05, "loss": 0.0395, "step": 21920 }, { "grad_norm": 0.4739632308483124, "learning_rate": 8.688542494514955e-05, "loss": 0.0383, "step": 21930 }, { "grad_norm": 0.6585289835929871, "learning_rate": 8.68714681753159e-05, "loss": 0.045, "step": 21940 }, { "grad_norm": 0.3633272349834442, "learning_rate": 8.685750510515676e-05, "loss": 0.0333, "step": 21950 }, { "grad_norm": 0.4251564145088196, "learning_rate": 8.684353573705805e-05, "loss": 0.039, "step": 21960 }, { "grad_norm": 0.3411434590816498, "learning_rate": 8.682956007340677e-05, "loss": 0.0276, "step": 21970 }, { "grad_norm": 0.42923182249069214, "learning_rate": 8.681557811659095e-05, "loss": 0.0319, "step": 21980 }, { "grad_norm": 0.32448166608810425, "learning_rate": 8.680158986899974e-05, "loss": 0.0278, "step": 21990 }, { "grad_norm": 0.3277810513973236, "learning_rate": 8.678759533302335e-05, "loss": 0.0312, "step": 22000 }, { "grad_norm": 0.36305999755859375, "learning_rate": 8.677359451105308e-05, "loss": 0.0283, "step": 22010 }, { "grad_norm": 0.30870726704597473, "learning_rate": 8.675958740548123e-05, "loss": 0.0295, "step": 22020 }, { "grad_norm": 0.38628292083740234, "learning_rate": 8.674557401870129e-05, "loss": 0.0401, "step": 22030 }, { "grad_norm": 0.38363131880760193, "learning_rate": 8.673155435310775e-05, "loss": 0.0415, "step": 22040 }, { "grad_norm": 0.2868172228336334, "learning_rate": 8.671752841109617e-05, "loss": 0.028, "step": 22050 }, { "grad_norm": 0.35808777809143066, "learning_rate": 8.670349619506321e-05, "loss": 0.036, "step": 22060 }, { "grad_norm": 0.4616454541683197, "learning_rate": 8.66894577074066e-05, "loss": 0.0365, "step": 22070 }, { "grad_norm": 0.38013872504234314, "learning_rate": 8.667541295052513e-05, "loss": 0.0326, "step": 22080 }, { "grad_norm": 0.41301316022872925, "learning_rate": 8.666136192681865e-05, "loss": 0.0412, "step": 22090 }, { "grad_norm": 0.30258017778396606, "learning_rate": 8.664730463868811e-05, "loss": 0.0394, "step": 22100 }, { "grad_norm": 0.4205940365791321, "learning_rate": 8.663324108853552e-05, "loss": 0.0442, "step": 22110 }, { "grad_norm": 0.4408998191356659, "learning_rate": 8.661917127876395e-05, "loss": 0.0566, "step": 22120 }, { "grad_norm": 0.29573243856430054, "learning_rate": 8.660509521177754e-05, "loss": 0.0446, "step": 22130 }, { "grad_norm": 0.48475587368011475, "learning_rate": 8.65910128899815e-05, "loss": 0.0353, "step": 22140 }, { "grad_norm": 0.43446195125579834, "learning_rate": 8.657692431578214e-05, "loss": 0.0354, "step": 22150 }, { "grad_norm": 0.46962493658065796, "learning_rate": 8.656282949158679e-05, "loss": 0.0374, "step": 22160 }, { "grad_norm": 0.2826853394508362, "learning_rate": 8.654872841980388e-05, "loss": 0.0271, "step": 22170 }, { "grad_norm": 0.3214128315448761, "learning_rate": 8.653462110284289e-05, "loss": 0.0286, "step": 22180 }, { "grad_norm": 0.3183860778808594, "learning_rate": 8.652050754311437e-05, "loss": 0.0351, "step": 22190 }, { "grad_norm": 0.31885451078414917, "learning_rate": 8.650638774302995e-05, "loss": 0.0321, "step": 22200 }, { "grad_norm": 0.32835495471954346, "learning_rate": 8.649226170500233e-05, "loss": 0.0369, "step": 22210 }, { "grad_norm": 0.29045674204826355, "learning_rate": 8.647812943144524e-05, "loss": 0.0345, "step": 22220 }, { "grad_norm": 0.28353890776634216, "learning_rate": 8.646399092477351e-05, "loss": 0.0318, "step": 22230 }, { "grad_norm": 0.4465060234069824, "learning_rate": 8.644984618740301e-05, "loss": 0.0453, "step": 22240 }, { "grad_norm": 0.3951113522052765, "learning_rate": 8.643569522175073e-05, "loss": 0.0287, "step": 22250 }, { "grad_norm": 0.4440721571445465, "learning_rate": 8.642153803023463e-05, "loss": 0.0364, "step": 22260 }, { "grad_norm": 0.5517642498016357, "learning_rate": 8.640737461527383e-05, "loss": 0.0325, "step": 22270 }, { "grad_norm": 0.37418192625045776, "learning_rate": 8.639320497928845e-05, "loss": 0.0392, "step": 22280 }, { "grad_norm": 0.37886038422584534, "learning_rate": 8.637902912469969e-05, "loss": 0.0298, "step": 22290 }, { "grad_norm": 0.3153965473175049, "learning_rate": 8.636484705392982e-05, "loss": 0.0325, "step": 22300 }, { "grad_norm": 0.4317663609981537, "learning_rate": 8.635065876940216e-05, "loss": 0.0339, "step": 22310 }, { "grad_norm": 0.3853519856929779, "learning_rate": 8.633646427354112e-05, "loss": 0.0376, "step": 22320 }, { "grad_norm": 0.3952181935310364, "learning_rate": 8.632226356877213e-05, "loss": 0.0421, "step": 22330 }, { "grad_norm": 0.26588475704193115, "learning_rate": 8.630805665752173e-05, "loss": 0.0441, "step": 22340 }, { "grad_norm": 0.27561813592910767, "learning_rate": 8.629384354221748e-05, "loss": 0.0401, "step": 22350 }, { "grad_norm": 0.3772193193435669, "learning_rate": 8.627962422528797e-05, "loss": 0.0342, "step": 22360 }, { "grad_norm": 0.34928277134895325, "learning_rate": 8.626539870916296e-05, "loss": 0.0505, "step": 22370 }, { "grad_norm": 0.31605809926986694, "learning_rate": 8.625116699627317e-05, "loss": 0.0342, "step": 22380 }, { "grad_norm": 0.3755585253238678, "learning_rate": 8.623692908905041e-05, "loss": 0.0318, "step": 22390 }, { "grad_norm": 0.5196133852005005, "learning_rate": 8.622268498992755e-05, "loss": 0.0445, "step": 22400 }, { "grad_norm": 0.4222351312637329, "learning_rate": 8.620843470133851e-05, "loss": 0.0588, "step": 22410 }, { "grad_norm": 0.4420446753501892, "learning_rate": 8.619417822571829e-05, "loss": 0.0424, "step": 22420 }, { "grad_norm": 0.3530055582523346, "learning_rate": 8.617991556550292e-05, "loss": 0.0351, "step": 22430 }, { "grad_norm": 0.5070646405220032, "learning_rate": 8.616564672312952e-05, "loss": 0.0368, "step": 22440 }, { "grad_norm": 0.4560467600822449, "learning_rate": 8.61513717010362e-05, "loss": 0.0429, "step": 22450 }, { "grad_norm": 0.52357017993927, "learning_rate": 8.613709050166221e-05, "loss": 0.0436, "step": 22460 }, { "grad_norm": 0.33340218663215637, "learning_rate": 8.61228031274478e-05, "loss": 0.0407, "step": 22470 }, { "grad_norm": 0.4015481770038605, "learning_rate": 8.610850958083431e-05, "loss": 0.0364, "step": 22480 }, { "grad_norm": 0.539232075214386, "learning_rate": 8.609420986426409e-05, "loss": 0.0359, "step": 22490 }, { "grad_norm": 0.43023762106895447, "learning_rate": 8.60799039801806e-05, "loss": 0.0351, "step": 22500 }, { "grad_norm": 0.6663268804550171, "learning_rate": 8.606559193102828e-05, "loss": 0.0408, "step": 22510 }, { "grad_norm": 0.5087714791297913, "learning_rate": 8.605127371925273e-05, "loss": 0.0491, "step": 22520 }, { "grad_norm": 0.4323459565639496, "learning_rate": 8.603694934730047e-05, "loss": 0.0467, "step": 22530 }, { "grad_norm": 0.2902677357196808, "learning_rate": 8.602261881761919e-05, "loss": 0.0284, "step": 22540 }, { "grad_norm": 0.3648397922515869, "learning_rate": 8.600828213265759e-05, "loss": 0.0505, "step": 22550 }, { "grad_norm": 0.41005924344062805, "learning_rate": 8.599393929486539e-05, "loss": 0.0466, "step": 22560 }, { "grad_norm": 0.41721582412719727, "learning_rate": 8.59795903066934e-05, "loss": 0.0406, "step": 22570 }, { "grad_norm": 0.3235006034374237, "learning_rate": 8.596523517059347e-05, "loss": 0.0379, "step": 22580 }, { "grad_norm": 0.3586238622665405, "learning_rate": 8.59508738890185e-05, "loss": 0.0317, "step": 22590 }, { "grad_norm": 0.4366118311882019, "learning_rate": 8.593650646442246e-05, "loss": 0.0398, "step": 22600 }, { "grad_norm": 0.3697560429573059, "learning_rate": 8.59221328992603e-05, "loss": 0.0356, "step": 22610 }, { "grad_norm": 0.4222295880317688, "learning_rate": 8.590775319598813e-05, "loss": 0.043, "step": 22620 }, { "grad_norm": 0.5075736045837402, "learning_rate": 8.589336735706301e-05, "loss": 0.046, "step": 22630 }, { "grad_norm": 0.38224703073501587, "learning_rate": 8.587897538494307e-05, "loss": 0.0561, "step": 22640 }, { "grad_norm": 0.45034658908843994, "learning_rate": 8.586457728208756e-05, "loss": 0.0427, "step": 22650 }, { "grad_norm": 0.47687751054763794, "learning_rate": 8.585017305095667e-05, "loss": 0.0406, "step": 22660 }, { "grad_norm": 0.4543975293636322, "learning_rate": 8.583576269401173e-05, "loss": 0.0391, "step": 22670 }, { "grad_norm": 0.6833778619766235, "learning_rate": 8.582134621371504e-05, "loss": 0.0454, "step": 22680 }, { "grad_norm": 0.6613524556159973, "learning_rate": 8.580692361253e-05, "loss": 0.0342, "step": 22690 }, { "grad_norm": 0.36710506677627563, "learning_rate": 8.579249489292104e-05, "loss": 0.0405, "step": 22700 }, { "grad_norm": 0.33938029408454895, "learning_rate": 8.577806005735363e-05, "loss": 0.0429, "step": 22710 }, { "grad_norm": 0.3935139775276184, "learning_rate": 8.576361910829429e-05, "loss": 0.0328, "step": 22720 }, { "grad_norm": 0.3049461543560028, "learning_rate": 8.574917204821057e-05, "loss": 0.0358, "step": 22730 }, { "grad_norm": 0.36120685935020447, "learning_rate": 8.57347188795711e-05, "loss": 0.0383, "step": 22740 }, { "grad_norm": 0.4500311315059662, "learning_rate": 8.572025960484551e-05, "loss": 0.0404, "step": 22750 }, { "grad_norm": 0.5113449096679688, "learning_rate": 8.57057942265045e-05, "loss": 0.0492, "step": 22760 }, { "grad_norm": 0.3937411606311798, "learning_rate": 8.569132274701984e-05, "loss": 0.0387, "step": 22770 }, { "grad_norm": 0.32340186834335327, "learning_rate": 8.567684516886427e-05, "loss": 0.0311, "step": 22780 }, { "grad_norm": 0.27151423692703247, "learning_rate": 8.56623614945116e-05, "loss": 0.0295, "step": 22790 }, { "grad_norm": 0.3852236270904541, "learning_rate": 8.564787172643675e-05, "loss": 0.0341, "step": 22800 }, { "grad_norm": 0.39577850699424744, "learning_rate": 8.563337586711559e-05, "loss": 0.0291, "step": 22810 }, { "grad_norm": 0.33994999527931213, "learning_rate": 8.561887391902506e-05, "loss": 0.0465, "step": 22820 }, { "grad_norm": 0.28700974583625793, "learning_rate": 8.560436588464316e-05, "loss": 0.032, "step": 22830 }, { "grad_norm": 0.29775652289390564, "learning_rate": 8.55898517664489e-05, "loss": 0.0445, "step": 22840 }, { "grad_norm": 0.4053298532962799, "learning_rate": 8.557533156692236e-05, "loss": 0.0481, "step": 22850 }, { "grad_norm": 0.34010130167007446, "learning_rate": 8.556080528854467e-05, "loss": 0.0328, "step": 22860 }, { "grad_norm": 0.41095197200775146, "learning_rate": 8.554627293379791e-05, "loss": 0.0328, "step": 22870 }, { "grad_norm": 0.4399034082889557, "learning_rate": 8.553173450516531e-05, "loss": 0.0435, "step": 22880 }, { "grad_norm": 0.38924890756607056, "learning_rate": 8.551719000513108e-05, "loss": 0.0391, "step": 22890 }, { "grad_norm": 0.44695156812667847, "learning_rate": 8.550263943618049e-05, "loss": 0.0548, "step": 22900 }, { "grad_norm": 0.4182230532169342, "learning_rate": 8.54880828007998e-05, "loss": 0.0388, "step": 22910 }, { "grad_norm": 0.4156601130962372, "learning_rate": 8.547352010147637e-05, "loss": 0.0457, "step": 22920 }, { "grad_norm": 0.312637060880661, "learning_rate": 8.545895134069855e-05, "loss": 0.0285, "step": 22930 }, { "grad_norm": 0.4367333948612213, "learning_rate": 8.544437652095576e-05, "loss": 0.0355, "step": 22940 }, { "grad_norm": 0.5285866260528564, "learning_rate": 8.542979564473843e-05, "loss": 0.0513, "step": 22950 }, { "grad_norm": 0.39202430844306946, "learning_rate": 8.541520871453802e-05, "loss": 0.0316, "step": 22960 }, { "grad_norm": 0.30243775248527527, "learning_rate": 8.540061573284705e-05, "loss": 0.0368, "step": 22970 }, { "grad_norm": 0.2559897005558014, "learning_rate": 8.538601670215906e-05, "loss": 0.0346, "step": 22980 }, { "grad_norm": 0.41328731179237366, "learning_rate": 8.537141162496864e-05, "loss": 0.0458, "step": 22990 }, { "grad_norm": 0.604231059551239, "learning_rate": 8.535680050377137e-05, "loss": 0.0333, "step": 23000 }, { "grad_norm": 0.3179367184638977, "learning_rate": 8.534218334106391e-05, "loss": 0.0355, "step": 23010 }, { "grad_norm": 0.3561067283153534, "learning_rate": 8.532756013934393e-05, "loss": 0.0338, "step": 23020 }, { "grad_norm": 0.38563087582588196, "learning_rate": 8.531293090111012e-05, "loss": 0.039, "step": 23030 }, { "grad_norm": 0.480266273021698, "learning_rate": 8.529829562886225e-05, "loss": 0.0367, "step": 23040 }, { "grad_norm": 0.43938249349594116, "learning_rate": 8.528365432510105e-05, "loss": 0.0496, "step": 23050 }, { "grad_norm": 0.3222350478172302, "learning_rate": 8.526900699232833e-05, "loss": 0.0417, "step": 23060 }, { "grad_norm": 0.505304217338562, "learning_rate": 8.525435363304695e-05, "loss": 0.038, "step": 23070 }, { "grad_norm": 0.4393477737903595, "learning_rate": 8.523969424976072e-05, "loss": 0.0364, "step": 23080 }, { "grad_norm": 0.3649662137031555, "learning_rate": 8.522502884497457e-05, "loss": 0.0371, "step": 23090 }, { "grad_norm": 0.5525692105293274, "learning_rate": 8.521035742119437e-05, "loss": 0.0444, "step": 23100 }, { "grad_norm": 0.3517846167087555, "learning_rate": 8.519567998092712e-05, "loss": 0.0435, "step": 23110 }, { "grad_norm": 0.4251403510570526, "learning_rate": 8.518099652668075e-05, "loss": 0.038, "step": 23120 }, { "grad_norm": 0.3180234730243683, "learning_rate": 8.516630706096429e-05, "loss": 0.0405, "step": 23130 }, { "grad_norm": 0.4311625361442566, "learning_rate": 8.515161158628773e-05, "loss": 0.0313, "step": 23140 }, { "grad_norm": 0.3357151448726654, "learning_rate": 8.513691010516216e-05, "loss": 0.0531, "step": 23150 }, { "grad_norm": 0.3122571110725403, "learning_rate": 8.512220262009966e-05, "loss": 0.0307, "step": 23160 }, { "grad_norm": 0.40387263894081116, "learning_rate": 8.510748913361332e-05, "loss": 0.0409, "step": 23170 }, { "grad_norm": 0.30973020195961, "learning_rate": 8.509276964821726e-05, "loss": 0.031, "step": 23180 }, { "grad_norm": 0.3961513936519623, "learning_rate": 8.507804416642669e-05, "loss": 0.0392, "step": 23190 }, { "grad_norm": 0.3277438282966614, "learning_rate": 8.506331269075774e-05, "loss": 0.0323, "step": 23200 }, { "grad_norm": 0.33613908290863037, "learning_rate": 8.504857522372765e-05, "loss": 0.0339, "step": 23210 }, { "grad_norm": 0.3684924840927124, "learning_rate": 8.503383176785461e-05, "loss": 0.0375, "step": 23220 }, { "grad_norm": 0.46014130115509033, "learning_rate": 8.501908232565792e-05, "loss": 0.0356, "step": 23230 }, { "grad_norm": 0.4170806109905243, "learning_rate": 8.50043268996578e-05, "loss": 0.0389, "step": 23240 }, { "grad_norm": 0.3277586102485657, "learning_rate": 8.498956549237562e-05, "loss": 0.0375, "step": 23250 }, { "grad_norm": 0.3323962986469269, "learning_rate": 8.497479810633366e-05, "loss": 0.0299, "step": 23260 }, { "grad_norm": 0.5002087950706482, "learning_rate": 8.496002474405525e-05, "loss": 0.0417, "step": 23270 }, { "grad_norm": 0.2705806791782379, "learning_rate": 8.494524540806478e-05, "loss": 0.042, "step": 23280 }, { "grad_norm": 0.2741166651248932, "learning_rate": 8.493046010088761e-05, "loss": 0.0561, "step": 23290 }, { "grad_norm": 3.0235962867736816, "learning_rate": 8.491566882505018e-05, "loss": 0.0465, "step": 23300 }, { "grad_norm": 0.4410336911678314, "learning_rate": 8.490087158307988e-05, "loss": 0.0558, "step": 23310 }, { "grad_norm": 0.6379101872444153, "learning_rate": 8.488606837750518e-05, "loss": 0.0495, "step": 23320 }, { "grad_norm": 0.4127291738986969, "learning_rate": 8.487125921085552e-05, "loss": 0.0384, "step": 23330 }, { "grad_norm": 0.3857043981552124, "learning_rate": 8.485644408566141e-05, "loss": 0.0403, "step": 23340 }, { "grad_norm": 0.34935081005096436, "learning_rate": 8.484162300445431e-05, "loss": 0.0354, "step": 23350 }, { "grad_norm": 0.4877939224243164, "learning_rate": 8.482679596976676e-05, "loss": 0.0297, "step": 23360 }, { "grad_norm": 0.3671572804450989, "learning_rate": 8.48119629841323e-05, "loss": 0.0348, "step": 23370 }, { "grad_norm": 0.37730884552001953, "learning_rate": 8.479712405008547e-05, "loss": 0.0385, "step": 23380 }, { "grad_norm": 0.26396891474723816, "learning_rate": 8.478227917016184e-05, "loss": 0.0441, "step": 23390 }, { "grad_norm": 0.2919887602329254, "learning_rate": 8.476742834689801e-05, "loss": 0.0408, "step": 23400 }, { "grad_norm": 0.4483569860458374, "learning_rate": 8.475257158283157e-05, "loss": 0.0387, "step": 23410 }, { "grad_norm": 0.38657882809638977, "learning_rate": 8.473770888050112e-05, "loss": 0.0314, "step": 23420 }, { "grad_norm": 0.33580613136291504, "learning_rate": 8.47228402424463e-05, "loss": 0.0316, "step": 23430 }, { "grad_norm": 1.147925853729248, "learning_rate": 8.470796567120775e-05, "loss": 0.052, "step": 23440 }, { "grad_norm": 0.4096258878707886, "learning_rate": 8.469308516932714e-05, "loss": 0.0425, "step": 23450 }, { "grad_norm": 0.4135271906852722, "learning_rate": 8.467819873934714e-05, "loss": 0.0394, "step": 23460 }, { "grad_norm": 0.4259633719921112, "learning_rate": 8.466330638381143e-05, "loss": 0.0404, "step": 23470 }, { "grad_norm": 0.33576899766921997, "learning_rate": 8.464840810526469e-05, "loss": 0.0368, "step": 23480 }, { "grad_norm": 0.40282002091407776, "learning_rate": 8.463350390625264e-05, "loss": 0.0321, "step": 23490 }, { "grad_norm": 0.32674795389175415, "learning_rate": 8.4618593789322e-05, "loss": 0.0436, "step": 23500 }, { "grad_norm": 0.3534316420555115, "learning_rate": 8.46036777570205e-05, "loss": 0.0391, "step": 23510 }, { "grad_norm": 0.5744674801826477, "learning_rate": 8.458875581189688e-05, "loss": 0.0614, "step": 23520 }, { "grad_norm": 0.33806660771369934, "learning_rate": 8.457382795650092e-05, "loss": 0.0319, "step": 23530 }, { "grad_norm": 0.34131643176078796, "learning_rate": 8.455889419338335e-05, "loss": 0.0412, "step": 23540 }, { "grad_norm": 0.42618128657341003, "learning_rate": 8.454395452509593e-05, "loss": 0.0339, "step": 23550 }, { "grad_norm": 0.3754461705684662, "learning_rate": 8.452900895419146e-05, "loss": 0.0353, "step": 23560 }, { "grad_norm": 0.23700550198554993, "learning_rate": 8.451405748322376e-05, "loss": 0.0308, "step": 23570 }, { "grad_norm": 0.3773941397666931, "learning_rate": 8.449910011474759e-05, "loss": 0.031, "step": 23580 }, { "grad_norm": 0.28100088238716125, "learning_rate": 8.448413685131876e-05, "loss": 0.0466, "step": 23590 }, { "grad_norm": 0.4102310538291931, "learning_rate": 8.446916769549407e-05, "loss": 0.0429, "step": 23600 }, { "grad_norm": 0.2893413007259369, "learning_rate": 8.445419264983136e-05, "loss": 0.0369, "step": 23610 }, { "grad_norm": 0.4247240126132965, "learning_rate": 8.443921171688947e-05, "loss": 0.0343, "step": 23620 }, { "grad_norm": 0.35962387919425964, "learning_rate": 8.442422489922819e-05, "loss": 0.0332, "step": 23630 }, { "grad_norm": 0.37285885214805603, "learning_rate": 8.440923219940838e-05, "loss": 0.0413, "step": 23640 }, { "grad_norm": 0.40125003457069397, "learning_rate": 8.439423361999189e-05, "loss": 0.0383, "step": 23650 }, { "grad_norm": 0.33894431591033936, "learning_rate": 8.437922916354155e-05, "loss": 0.0421, "step": 23660 }, { "grad_norm": 0.34657227993011475, "learning_rate": 8.436421883262123e-05, "loss": 0.0312, "step": 23670 }, { "grad_norm": 0.3013048768043518, "learning_rate": 8.434920262979577e-05, "loss": 0.0521, "step": 23680 }, { "grad_norm": 0.532938539981842, "learning_rate": 8.433418055763104e-05, "loss": 0.0284, "step": 23690 }, { "grad_norm": 0.35082700848579407, "learning_rate": 8.431915261869389e-05, "loss": 0.0382, "step": 23700 }, { "grad_norm": 0.4770638048648834, "learning_rate": 8.43041188155522e-05, "loss": 0.0408, "step": 23710 }, { "grad_norm": 0.48580142855644226, "learning_rate": 8.428907915077481e-05, "loss": 0.0378, "step": 23720 }, { "grad_norm": 0.4573436677455902, "learning_rate": 8.42740336269316e-05, "loss": 0.0449, "step": 23730 }, { "grad_norm": 0.17771008610725403, "learning_rate": 8.425898224659345e-05, "loss": 0.0298, "step": 23740 }, { "grad_norm": 0.436710387468338, "learning_rate": 8.42439250123322e-05, "loss": 0.0319, "step": 23750 }, { "grad_norm": 0.45633408427238464, "learning_rate": 8.422886192672076e-05, "loss": 0.0323, "step": 23760 }, { "grad_norm": 0.4004127085208893, "learning_rate": 8.421379299233297e-05, "loss": 0.0326, "step": 23770 }, { "grad_norm": 0.35220882296562195, "learning_rate": 8.419871821174371e-05, "loss": 0.0325, "step": 23780 }, { "grad_norm": 0.5530247092247009, "learning_rate": 8.418363758752884e-05, "loss": 0.0309, "step": 23790 }, { "grad_norm": 0.49172109365463257, "learning_rate": 8.416855112226523e-05, "loss": 0.0382, "step": 23800 }, { "grad_norm": 0.3813387453556061, "learning_rate": 8.415345881853075e-05, "loss": 0.0327, "step": 23810 }, { "grad_norm": 0.3101601302623749, "learning_rate": 8.413836067890426e-05, "loss": 0.0297, "step": 23820 }, { "grad_norm": 0.2953268885612488, "learning_rate": 8.41232567059656e-05, "loss": 0.0298, "step": 23830 }, { "grad_norm": 0.38912081718444824, "learning_rate": 8.410814690229565e-05, "loss": 0.0317, "step": 23840 }, { "grad_norm": 0.2035950869321823, "learning_rate": 8.409303127047626e-05, "loss": 0.0325, "step": 23850 }, { "grad_norm": 0.4077439606189728, "learning_rate": 8.407790981309028e-05, "loss": 0.0317, "step": 23860 }, { "grad_norm": 0.385945200920105, "learning_rate": 8.406278253272153e-05, "loss": 0.0378, "step": 23870 }, { "grad_norm": 0.5022242665290833, "learning_rate": 8.404764943195487e-05, "loss": 0.0312, "step": 23880 }, { "grad_norm": 0.32013094425201416, "learning_rate": 8.403251051337613e-05, "loss": 0.0487, "step": 23890 }, { "grad_norm": 0.35828423500061035, "learning_rate": 8.401736577957214e-05, "loss": 0.0364, "step": 23900 }, { "grad_norm": 0.2262554168701172, "learning_rate": 8.40022152331307e-05, "loss": 0.0368, "step": 23910 }, { "grad_norm": 0.36778074502944946, "learning_rate": 8.398705887664064e-05, "loss": 0.0454, "step": 23920 }, { "grad_norm": 0.24582016468048096, "learning_rate": 8.397189671269177e-05, "loss": 0.0317, "step": 23930 }, { "grad_norm": 0.4018858075141907, "learning_rate": 8.395672874387488e-05, "loss": 0.0276, "step": 23940 }, { "grad_norm": 0.24715685844421387, "learning_rate": 8.394155497278177e-05, "loss": 0.0333, "step": 23950 }, { "grad_norm": 0.4592295289039612, "learning_rate": 8.392637540200523e-05, "loss": 0.0475, "step": 23960 }, { "grad_norm": 0.39657095074653625, "learning_rate": 8.391119003413902e-05, "loss": 0.0379, "step": 23970 }, { "grad_norm": 0.3775648772716522, "learning_rate": 8.38959988717779e-05, "loss": 0.0348, "step": 23980 }, { "grad_norm": 0.4042321741580963, "learning_rate": 8.388080191751764e-05, "loss": 0.0408, "step": 23990 }, { "grad_norm": 0.45783811807632446, "learning_rate": 8.386559917395496e-05, "loss": 0.0372, "step": 24000 }, { "grad_norm": 0.39051398634910583, "learning_rate": 8.385039064368761e-05, "loss": 0.0275, "step": 24010 }, { "grad_norm": 0.47246918082237244, "learning_rate": 8.383517632931431e-05, "loss": 0.0279, "step": 24020 }, { "grad_norm": 0.3614782392978668, "learning_rate": 8.381995623343477e-05, "loss": 0.034, "step": 24030 }, { "grad_norm": 0.5248852372169495, "learning_rate": 8.380473035864968e-05, "loss": 0.0374, "step": 24040 }, { "grad_norm": 0.37915700674057007, "learning_rate": 8.378949870756076e-05, "loss": 0.0323, "step": 24050 }, { "grad_norm": 0.4802532196044922, "learning_rate": 8.377426128277063e-05, "loss": 0.0383, "step": 24060 }, { "grad_norm": 0.4315812289714813, "learning_rate": 8.375901808688298e-05, "loss": 0.0283, "step": 24070 }, { "grad_norm": 0.3957928717136383, "learning_rate": 8.374376912250246e-05, "loss": 0.0415, "step": 24080 }, { "grad_norm": 0.4168495535850525, "learning_rate": 8.372851439223468e-05, "loss": 0.0315, "step": 24090 }, { "grad_norm": 0.30233821272850037, "learning_rate": 8.371325389868627e-05, "loss": 0.0391, "step": 24100 }, { "grad_norm": 0.4667510390281677, "learning_rate": 8.369798764446482e-05, "loss": 0.0316, "step": 24110 }, { "grad_norm": 0.3848114311695099, "learning_rate": 8.368271563217893e-05, "loss": 0.0342, "step": 24120 }, { "grad_norm": 0.34600669145584106, "learning_rate": 8.366743786443817e-05, "loss": 0.0389, "step": 24130 }, { "grad_norm": 0.32399314641952515, "learning_rate": 8.365215434385309e-05, "loss": 0.0423, "step": 24140 }, { "grad_norm": 0.41041192412376404, "learning_rate": 8.36368650730352e-05, "loss": 0.0384, "step": 24150 }, { "grad_norm": 0.29271399974823, "learning_rate": 8.362157005459705e-05, "loss": 0.0399, "step": 24160 }, { "grad_norm": 0.3455059230327606, "learning_rate": 8.360626929115213e-05, "loss": 0.0279, "step": 24170 }, { "grad_norm": 0.4523559808731079, "learning_rate": 8.359096278531492e-05, "loss": 0.0304, "step": 24180 }, { "grad_norm": 0.3605392575263977, "learning_rate": 8.357565053970088e-05, "loss": 0.031, "step": 24190 }, { "grad_norm": 0.39659714698791504, "learning_rate": 8.356033255692647e-05, "loss": 0.0309, "step": 24200 }, { "grad_norm": 0.4788832366466522, "learning_rate": 8.354500883960911e-05, "loss": 0.0352, "step": 24210 }, { "grad_norm": 0.4099944829940796, "learning_rate": 8.352967939036717e-05, "loss": 0.0347, "step": 24220 }, { "grad_norm": 0.33487409353256226, "learning_rate": 8.35143442118201e-05, "loss": 0.0401, "step": 24230 }, { "grad_norm": 0.21711169183254242, "learning_rate": 8.349900330658819e-05, "loss": 0.0288, "step": 24240 }, { "grad_norm": 0.36788222193717957, "learning_rate": 8.348365667729284e-05, "loss": 0.0379, "step": 24250 }, { "grad_norm": 0.3923354744911194, "learning_rate": 8.346830432655633e-05, "loss": 0.0291, "step": 24260 }, { "grad_norm": 0.5786794424057007, "learning_rate": 8.345294625700195e-05, "loss": 0.028, "step": 24270 }, { "grad_norm": 0.4410168528556824, "learning_rate": 8.343758247125402e-05, "loss": 0.0331, "step": 24280 }, { "grad_norm": 0.36910557746887207, "learning_rate": 8.342221297193776e-05, "loss": 0.0359, "step": 24290 }, { "grad_norm": 0.37725257873535156, "learning_rate": 8.34068377616794e-05, "loss": 0.0419, "step": 24300 }, { "grad_norm": 0.5156923532485962, "learning_rate": 8.339145684310615e-05, "loss": 0.0405, "step": 24310 }, { "grad_norm": 0.448604941368103, "learning_rate": 8.337607021884618e-05, "loss": 0.0325, "step": 24320 }, { "grad_norm": 0.22362004220485687, "learning_rate": 8.336067789152867e-05, "loss": 0.0259, "step": 24330 }, { "grad_norm": 0.3250076472759247, "learning_rate": 8.334527986378369e-05, "loss": 0.0361, "step": 24340 }, { "grad_norm": 0.31865835189819336, "learning_rate": 8.332987613824239e-05, "loss": 0.0393, "step": 24350 }, { "grad_norm": 0.3206695318222046, "learning_rate": 8.331446671753685e-05, "loss": 0.0343, "step": 24360 }, { "grad_norm": 0.36931493878364563, "learning_rate": 8.329905160430007e-05, "loss": 0.0341, "step": 24370 }, { "grad_norm": 0.42978498339653015, "learning_rate": 8.328363080116611e-05, "loss": 0.0302, "step": 24380 }, { "grad_norm": 0.2510172128677368, "learning_rate": 8.326820431076997e-05, "loss": 0.0301, "step": 24390 }, { "grad_norm": 0.31825393438339233, "learning_rate": 8.325277213574759e-05, "loss": 0.0344, "step": 24400 }, { "grad_norm": 0.3764890730381012, "learning_rate": 8.32373342787359e-05, "loss": 0.0356, "step": 24410 }, { "grad_norm": 0.33550673723220825, "learning_rate": 8.322189074237285e-05, "loss": 0.0339, "step": 24420 }, { "grad_norm": 0.3194847106933594, "learning_rate": 8.32064415292973e-05, "loss": 0.0357, "step": 24430 }, { "grad_norm": 0.24910366535186768, "learning_rate": 8.319098664214907e-05, "loss": 0.0323, "step": 24440 }, { "grad_norm": 0.5000767707824707, "learning_rate": 8.3175526083569e-05, "loss": 0.0405, "step": 24450 }, { "grad_norm": 0.29565975069999695, "learning_rate": 8.316005985619889e-05, "loss": 0.038, "step": 24460 }, { "grad_norm": 0.6509681344032288, "learning_rate": 8.314458796268147e-05, "loss": 0.0393, "step": 24470 }, { "grad_norm": 0.4527994692325592, "learning_rate": 8.312911040566047e-05, "loss": 0.0348, "step": 24480 }, { "grad_norm": 0.3472166061401367, "learning_rate": 8.31136271877806e-05, "loss": 0.0344, "step": 24490 }, { "grad_norm": 0.4252328872680664, "learning_rate": 8.309813831168748e-05, "loss": 0.0412, "step": 24500 }, { "grad_norm": 0.49496132135391235, "learning_rate": 8.308264378002777e-05, "loss": 0.0358, "step": 24510 }, { "grad_norm": 0.24624080955982208, "learning_rate": 8.306714359544906e-05, "loss": 0.0404, "step": 24520 }, { "grad_norm": 0.36535438895225525, "learning_rate": 8.30516377605999e-05, "loss": 0.0293, "step": 24530 }, { "grad_norm": 0.43907228112220764, "learning_rate": 8.30361262781298e-05, "loss": 0.0365, "step": 24540 }, { "grad_norm": 0.41134047508239746, "learning_rate": 8.302060915068924e-05, "loss": 0.0399, "step": 24550 }, { "grad_norm": 0.39012420177459717, "learning_rate": 8.300508638092972e-05, "loss": 0.0314, "step": 24560 }, { "grad_norm": 0.41651543974876404, "learning_rate": 8.298955797150361e-05, "loss": 0.0322, "step": 24570 }, { "grad_norm": 0.4485474228858948, "learning_rate": 8.297402392506433e-05, "loss": 0.0371, "step": 24580 }, { "grad_norm": 0.4116566479206085, "learning_rate": 8.295848424426617e-05, "loss": 0.0318, "step": 24590 }, { "grad_norm": 0.3045838177204132, "learning_rate": 8.29429389317645e-05, "loss": 0.0305, "step": 24600 }, { "grad_norm": 0.4169469177722931, "learning_rate": 8.292738799021556e-05, "loss": 0.0432, "step": 24610 }, { "grad_norm": 0.41946718096733093, "learning_rate": 8.291183142227656e-05, "loss": 0.0297, "step": 24620 }, { "grad_norm": 0.2616977393627167, "learning_rate": 8.289626923060572e-05, "loss": 0.0408, "step": 24630 }, { "grad_norm": 0.35977181792259216, "learning_rate": 8.288070141786218e-05, "loss": 0.0379, "step": 24640 }, { "grad_norm": 0.2868396043777466, "learning_rate": 8.286512798670605e-05, "loss": 0.0356, "step": 24650 }, { "grad_norm": 0.35753878951072693, "learning_rate": 8.284954893979842e-05, "loss": 0.0378, "step": 24660 }, { "grad_norm": 0.3017912805080414, "learning_rate": 8.283396427980131e-05, "loss": 0.0452, "step": 24670 }, { "grad_norm": 0.31488609313964844, "learning_rate": 8.281837400937771e-05, "loss": 0.0356, "step": 24680 }, { "grad_norm": 0.49932029843330383, "learning_rate": 8.28027781311916e-05, "loss": 0.0344, "step": 24690 }, { "grad_norm": 0.7879554033279419, "learning_rate": 8.278717664790785e-05, "loss": 0.0452, "step": 24700 }, { "grad_norm": 0.37553805112838745, "learning_rate": 8.277156956219234e-05, "loss": 0.0351, "step": 24710 }, { "grad_norm": 0.38068220019340515, "learning_rate": 8.275595687671189e-05, "loss": 0.0354, "step": 24720 }, { "grad_norm": 0.4475862979888916, "learning_rate": 8.27403385941343e-05, "loss": 0.0379, "step": 24730 }, { "grad_norm": 0.3700030744075775, "learning_rate": 8.272471471712828e-05, "loss": 0.0345, "step": 24740 }, { "grad_norm": 0.33229702711105347, "learning_rate": 8.270908524836355e-05, "loss": 0.039, "step": 24750 }, { "grad_norm": 0.33772534132003784, "learning_rate": 8.269345019051074e-05, "loss": 0.035, "step": 24760 }, { "grad_norm": 0.318027526140213, "learning_rate": 8.267780954624147e-05, "loss": 0.0342, "step": 24770 }, { "grad_norm": 0.4669472873210907, "learning_rate": 8.266216331822827e-05, "loss": 0.0409, "step": 24780 }, { "grad_norm": 0.43887314200401306, "learning_rate": 8.264651150914469e-05, "loss": 0.0363, "step": 24790 }, { "grad_norm": 0.3980471193790436, "learning_rate": 8.263085412166517e-05, "loss": 0.0324, "step": 24800 }, { "grad_norm": 0.42711034417152405, "learning_rate": 8.261519115846514e-05, "loss": 0.0341, "step": 24810 }, { "grad_norm": 0.37991347908973694, "learning_rate": 8.259952262222096e-05, "loss": 0.0313, "step": 24820 }, { "grad_norm": 0.4654092490673065, "learning_rate": 8.258384851560997e-05, "loss": 0.0358, "step": 24830 }, { "grad_norm": 0.6108965277671814, "learning_rate": 8.256816884131044e-05, "loss": 0.0369, "step": 24840 }, { "grad_norm": 0.5034791827201843, "learning_rate": 8.255248360200159e-05, "loss": 0.0376, "step": 24850 }, { "grad_norm": 0.3783958852291107, "learning_rate": 8.253679280036359e-05, "loss": 0.0304, "step": 24860 }, { "grad_norm": 0.5059083104133606, "learning_rate": 8.252109643907762e-05, "loss": 0.047, "step": 24870 }, { "grad_norm": 0.41055649518966675, "learning_rate": 8.250539452082569e-05, "loss": 0.0369, "step": 24880 }, { "grad_norm": 0.3796708285808563, "learning_rate": 8.248968704829087e-05, "loss": 0.0332, "step": 24890 }, { "grad_norm": 0.3615628778934479, "learning_rate": 8.247397402415714e-05, "loss": 0.0423, "step": 24900 }, { "grad_norm": 0.35557490587234497, "learning_rate": 8.24582554511094e-05, "loss": 0.0346, "step": 24910 }, { "grad_norm": 0.3410646319389343, "learning_rate": 8.244253133183355e-05, "loss": 0.0291, "step": 24920 }, { "grad_norm": 0.5461837649345398, "learning_rate": 8.24268016690164e-05, "loss": 0.0341, "step": 24930 }, { "grad_norm": 0.38868650794029236, "learning_rate": 8.241106646534571e-05, "loss": 0.0405, "step": 24940 }, { "grad_norm": 0.4281044006347656, "learning_rate": 8.23953257235102e-05, "loss": 0.0423, "step": 24950 }, { "grad_norm": 0.5085256099700928, "learning_rate": 8.237957944619956e-05, "loss": 0.0364, "step": 24960 }, { "grad_norm": 0.29354703426361084, "learning_rate": 8.236382763610437e-05, "loss": 0.0355, "step": 24970 }, { "grad_norm": 0.7171162366867065, "learning_rate": 8.234807029591619e-05, "loss": 0.0407, "step": 24980 }, { "grad_norm": 0.47157251834869385, "learning_rate": 8.233230742832752e-05, "loss": 0.0505, "step": 24990 }, { "grad_norm": 0.5246955752372742, "learning_rate": 8.231653903603178e-05, "loss": 0.0402, "step": 25000 }, { "grad_norm": 0.27632060647010803, "learning_rate": 8.23007651217234e-05, "loss": 0.037, "step": 25010 }, { "grad_norm": 0.4500003159046173, "learning_rate": 8.228498568809769e-05, "loss": 0.0351, "step": 25020 }, { "grad_norm": 0.47124117612838745, "learning_rate": 8.22692007378509e-05, "loss": 0.0463, "step": 25030 }, { "grad_norm": 0.469471275806427, "learning_rate": 8.225341027368028e-05, "loss": 0.0418, "step": 25040 }, { "grad_norm": 0.3079167306423187, "learning_rate": 8.223761429828399e-05, "loss": 0.0384, "step": 25050 }, { "grad_norm": 0.41714802384376526, "learning_rate": 8.22218128143611e-05, "loss": 0.0468, "step": 25060 }, { "grad_norm": 0.41352200508117676, "learning_rate": 8.220600582461166e-05, "loss": 0.0377, "step": 25070 }, { "grad_norm": 0.2438250631093979, "learning_rate": 8.219019333173668e-05, "loss": 0.0469, "step": 25080 }, { "grad_norm": 0.6904088258743286, "learning_rate": 8.217437533843805e-05, "loss": 0.0493, "step": 25090 }, { "grad_norm": 0.35696548223495483, "learning_rate": 8.215855184741867e-05, "loss": 0.0411, "step": 25100 }, { "grad_norm": 0.39176124334335327, "learning_rate": 8.21427228613823e-05, "loss": 0.037, "step": 25110 }, { "grad_norm": 0.4464782476425171, "learning_rate": 8.21268883830337e-05, "loss": 0.0406, "step": 25120 }, { "grad_norm": 0.4615381360054016, "learning_rate": 8.211104841507855e-05, "loss": 0.0483, "step": 25130 }, { "grad_norm": 0.46791747212409973, "learning_rate": 8.209520296022346e-05, "loss": 0.0388, "step": 25140 }, { "grad_norm": 0.47856998443603516, "learning_rate": 8.207935202117599e-05, "loss": 0.0402, "step": 25150 }, { "grad_norm": 0.27491575479507446, "learning_rate": 8.206349560064463e-05, "loss": 0.0367, "step": 25160 }, { "grad_norm": 0.3849489092826843, "learning_rate": 8.204763370133881e-05, "loss": 0.027, "step": 25170 }, { "grad_norm": 0.45738449692726135, "learning_rate": 8.203176632596892e-05, "loss": 0.0378, "step": 25180 }, { "grad_norm": 0.3607103228569031, "learning_rate": 8.20158934772462e-05, "loss": 0.0401, "step": 25190 }, { "grad_norm": 0.28482091426849365, "learning_rate": 8.200001515788294e-05, "loss": 0.0296, "step": 25200 }, { "grad_norm": 0.38196489214897156, "learning_rate": 8.198413137059228e-05, "loss": 0.0379, "step": 25210 }, { "grad_norm": 0.31615501642227173, "learning_rate": 8.196824211808835e-05, "loss": 0.0314, "step": 25220 }, { "grad_norm": 0.3601246476173401, "learning_rate": 8.195234740308617e-05, "loss": 0.0413, "step": 25230 }, { "grad_norm": 0.2889067828655243, "learning_rate": 8.193644722830171e-05, "loss": 0.0291, "step": 25240 }, { "grad_norm": 0.47116804122924805, "learning_rate": 8.19205415964519e-05, "loss": 0.0297, "step": 25250 }, { "grad_norm": 0.3750629425048828, "learning_rate": 8.190463051025456e-05, "loss": 0.0275, "step": 25260 }, { "grad_norm": 0.3221515417098999, "learning_rate": 8.188871397242843e-05, "loss": 0.0345, "step": 25270 }, { "grad_norm": 0.311146080493927, "learning_rate": 8.187279198569326e-05, "loss": 0.0315, "step": 25280 }, { "grad_norm": 0.2923312187194824, "learning_rate": 8.185686455276966e-05, "loss": 0.0291, "step": 25290 }, { "grad_norm": 0.37941667437553406, "learning_rate": 8.184093167637921e-05, "loss": 0.0338, "step": 25300 }, { "grad_norm": 0.43330952525138855, "learning_rate": 8.182499335924437e-05, "loss": 0.0314, "step": 25310 }, { "grad_norm": 0.48914027214050293, "learning_rate": 8.18090496040886e-05, "loss": 0.0306, "step": 25320 }, { "grad_norm": 0.32816556096076965, "learning_rate": 8.179310041363621e-05, "loss": 0.0406, "step": 25330 }, { "grad_norm": 0.34060803055763245, "learning_rate": 8.17771457906125e-05, "loss": 0.0368, "step": 25340 }, { "grad_norm": 0.3758247494697571, "learning_rate": 8.176118573774371e-05, "loss": 0.0327, "step": 25350 }, { "grad_norm": 0.3645665943622589, "learning_rate": 8.174522025775692e-05, "loss": 0.0313, "step": 25360 }, { "grad_norm": 0.4654335081577301, "learning_rate": 8.172924935338022e-05, "loss": 0.036, "step": 25370 }, { "grad_norm": 0.47432345151901245, "learning_rate": 8.171327302734262e-05, "loss": 0.0309, "step": 25380 }, { "grad_norm": 0.4040059447288513, "learning_rate": 8.169729128237401e-05, "loss": 0.0285, "step": 25390 }, { "grad_norm": 0.2895215153694153, "learning_rate": 8.168130412120525e-05, "loss": 0.0345, "step": 25400 }, { "grad_norm": 1.0266969203948975, "learning_rate": 8.16653115465681e-05, "loss": 0.0358, "step": 25410 }, { "grad_norm": 0.5577085018157959, "learning_rate": 8.164931356119526e-05, "loss": 0.0428, "step": 25420 }, { "grad_norm": 0.6250331997871399, "learning_rate": 8.163331016782032e-05, "loss": 0.0497, "step": 25430 }, { "grad_norm": 0.39714840054512024, "learning_rate": 8.161730136917785e-05, "loss": 0.0431, "step": 25440 }, { "grad_norm": 0.48478174209594727, "learning_rate": 8.160128716800333e-05, "loss": 0.05, "step": 25450 }, { "grad_norm": 0.43609562516212463, "learning_rate": 8.158526756703313e-05, "loss": 0.0318, "step": 25460 }, { "grad_norm": 0.42268604040145874, "learning_rate": 8.156924256900455e-05, "loss": 0.0378, "step": 25470 }, { "grad_norm": 0.40164485573768616, "learning_rate": 8.155321217665584e-05, "loss": 0.0325, "step": 25480 }, { "grad_norm": 0.4610258638858795, "learning_rate": 8.153717639272614e-05, "loss": 0.0321, "step": 25490 }, { "grad_norm": 0.28739556670188904, "learning_rate": 8.152113521995555e-05, "loss": 0.0345, "step": 25500 }, { "grad_norm": 0.4505804777145386, "learning_rate": 8.150508866108505e-05, "loss": 0.0324, "step": 25510 }, { "grad_norm": 0.33303236961364746, "learning_rate": 8.148903671885657e-05, "loss": 0.0326, "step": 25520 }, { "grad_norm": 0.4411095976829529, "learning_rate": 8.147297939601292e-05, "loss": 0.053, "step": 25530 }, { "grad_norm": 0.43765076994895935, "learning_rate": 8.145691669529792e-05, "loss": 0.0307, "step": 25540 }, { "grad_norm": 0.5842294692993164, "learning_rate": 8.144084861945618e-05, "loss": 0.0362, "step": 25550 }, { "grad_norm": 0.4445098638534546, "learning_rate": 8.142477517123333e-05, "loss": 0.0391, "step": 25560 }, { "grad_norm": 0.3652033805847168, "learning_rate": 8.140869635337586e-05, "loss": 0.0321, "step": 25570 }, { "grad_norm": 0.24020975828170776, "learning_rate": 8.139261216863123e-05, "loss": 0.0261, "step": 25580 }, { "grad_norm": 0.4195672571659088, "learning_rate": 8.137652261974776e-05, "loss": 0.0377, "step": 25590 }, { "grad_norm": 0.4788265824317932, "learning_rate": 8.136042770947472e-05, "loss": 0.0332, "step": 25600 }, { "grad_norm": 0.365705281496048, "learning_rate": 8.134432744056228e-05, "loss": 0.0437, "step": 25610 }, { "grad_norm": 0.47579529881477356, "learning_rate": 8.132822181576158e-05, "loss": 0.0431, "step": 25620 }, { "grad_norm": 0.3574933111667633, "learning_rate": 8.131211083782459e-05, "loss": 0.0333, "step": 25630 }, { "grad_norm": 0.25114819407463074, "learning_rate": 8.129599450950424e-05, "loss": 0.0328, "step": 25640 }, { "grad_norm": 0.5007825493812561, "learning_rate": 8.127987283355438e-05, "loss": 0.039, "step": 25650 }, { "grad_norm": 0.32527098059654236, "learning_rate": 8.126374581272976e-05, "loss": 0.0247, "step": 25660 }, { "grad_norm": 0.4138849973678589, "learning_rate": 8.124761344978605e-05, "loss": 0.0464, "step": 25670 }, { "grad_norm": 0.30735766887664795, "learning_rate": 8.12314757474798e-05, "loss": 0.0282, "step": 25680 }, { "grad_norm": 0.443234920501709, "learning_rate": 8.121533270856856e-05, "loss": 0.0308, "step": 25690 }, { "grad_norm": 0.40372034907341003, "learning_rate": 8.119918433581069e-05, "loss": 0.0327, "step": 25700 }, { "grad_norm": 0.43141043186187744, "learning_rate": 8.118303063196551e-05, "loss": 0.0375, "step": 25710 }, { "grad_norm": 0.29713883996009827, "learning_rate": 8.116687159979326e-05, "loss": 0.0332, "step": 25720 }, { "grad_norm": 0.416826069355011, "learning_rate": 8.115070724205508e-05, "loss": 0.0352, "step": 25730 }, { "grad_norm": 0.3783033490180969, "learning_rate": 8.113453756151296e-05, "loss": 0.0361, "step": 25740 }, { "grad_norm": 0.5108856558799744, "learning_rate": 8.111836256092995e-05, "loss": 0.0325, "step": 25750 }, { "grad_norm": 0.7016295790672302, "learning_rate": 8.110218224306985e-05, "loss": 0.0363, "step": 25760 }, { "grad_norm": 0.42586734890937805, "learning_rate": 8.108599661069745e-05, "loss": 0.0365, "step": 25770 }, { "grad_norm": 0.4369219243526459, "learning_rate": 8.106980566657845e-05, "loss": 0.0314, "step": 25780 }, { "grad_norm": 0.2927134037017822, "learning_rate": 8.10536094134794e-05, "loss": 0.0328, "step": 25790 }, { "grad_norm": 0.3858529329299927, "learning_rate": 8.103740785416783e-05, "loss": 0.03, "step": 25800 }, { "grad_norm": 0.40647292137145996, "learning_rate": 8.102120099141212e-05, "loss": 0.0367, "step": 25810 }, { "grad_norm": 0.30027517676353455, "learning_rate": 8.100498882798163e-05, "loss": 0.0271, "step": 25820 }, { "grad_norm": 0.27256646752357483, "learning_rate": 8.09887713666465e-05, "loss": 0.0256, "step": 25830 }, { "grad_norm": 0.3578132092952728, "learning_rate": 8.09725486101779e-05, "loss": 0.0339, "step": 25840 }, { "grad_norm": 0.30504029989242554, "learning_rate": 8.095632056134784e-05, "loss": 0.0398, "step": 25850 }, { "grad_norm": 0.3377775549888611, "learning_rate": 8.094008722292925e-05, "loss": 0.0325, "step": 25860 }, { "grad_norm": 0.37132880091667175, "learning_rate": 8.092384859769598e-05, "loss": 0.0331, "step": 25870 }, { "grad_norm": 0.3806513845920563, "learning_rate": 8.090760468842275e-05, "loss": 0.0412, "step": 25880 }, { "grad_norm": 0.3047715723514557, "learning_rate": 8.089135549788521e-05, "loss": 0.0356, "step": 25890 }, { "grad_norm": 0.4060424566268921, "learning_rate": 8.087510102885987e-05, "loss": 0.0251, "step": 25900 }, { "grad_norm": 0.4474785327911377, "learning_rate": 8.085884128412422e-05, "loss": 0.0331, "step": 25910 }, { "grad_norm": 0.43653708696365356, "learning_rate": 8.084257626645659e-05, "loss": 0.0331, "step": 25920 }, { "grad_norm": 0.5509287714958191, "learning_rate": 8.08263059786362e-05, "loss": 0.0352, "step": 25930 }, { "grad_norm": 0.4686545133590698, "learning_rate": 8.081003042344325e-05, "loss": 0.036, "step": 25940 }, { "grad_norm": 0.3334846496582031, "learning_rate": 8.079374960365872e-05, "loss": 0.036, "step": 25950 }, { "grad_norm": 0.32341787219047546, "learning_rate": 8.077746352206463e-05, "loss": 0.0375, "step": 25960 }, { "grad_norm": 1.0562037229537964, "learning_rate": 8.076117218144377e-05, "loss": 0.044, "step": 25970 }, { "grad_norm": 0.3957820236682892, "learning_rate": 8.074487558457991e-05, "loss": 0.0345, "step": 25980 }, { "grad_norm": 0.3159516751766205, "learning_rate": 8.072857373425768e-05, "loss": 0.0362, "step": 25990 }, { "grad_norm": 0.4394740164279938, "learning_rate": 8.071226663326264e-05, "loss": 0.0357, "step": 26000 }, { "grad_norm": 0.43020036816596985, "learning_rate": 8.069595428438121e-05, "loss": 0.0399, "step": 26010 }, { "grad_norm": 0.32200953364372253, "learning_rate": 8.067963669040072e-05, "loss": 0.0377, "step": 26020 }, { "grad_norm": 0.530143678188324, "learning_rate": 8.066331385410942e-05, "loss": 0.0398, "step": 26030 }, { "grad_norm": 0.4080372452735901, "learning_rate": 8.064698577829641e-05, "loss": 0.0406, "step": 26040 }, { "grad_norm": 0.3103410005569458, "learning_rate": 8.063065246575175e-05, "loss": 0.0385, "step": 26050 }, { "grad_norm": 0.5109865665435791, "learning_rate": 8.061431391926631e-05, "loss": 0.0362, "step": 26060 }, { "grad_norm": 0.3276873528957367, "learning_rate": 8.059797014163195e-05, "loss": 0.041, "step": 26070 }, { "grad_norm": 0.4645107090473175, "learning_rate": 8.058162113564133e-05, "loss": 0.0466, "step": 26080 }, { "grad_norm": 0.3627856373786926, "learning_rate": 8.056526690408806e-05, "loss": 0.0413, "step": 26090 }, { "grad_norm": 0.39860832691192627, "learning_rate": 8.054890744976666e-05, "loss": 0.0375, "step": 26100 }, { "grad_norm": 0.32583481073379517, "learning_rate": 8.053254277547248e-05, "loss": 0.0276, "step": 26110 }, { "grad_norm": 0.5081115961074829, "learning_rate": 8.051617288400182e-05, "loss": 0.0325, "step": 26120 }, { "grad_norm": 0.32567667961120605, "learning_rate": 8.049979777815182e-05, "loss": 0.0368, "step": 26130 }, { "grad_norm": 0.3688574731349945, "learning_rate": 8.048341746072054e-05, "loss": 0.0333, "step": 26140 }, { "grad_norm": 0.4307560324668884, "learning_rate": 8.046703193450696e-05, "loss": 0.0383, "step": 26150 }, { "grad_norm": 0.3876507878303528, "learning_rate": 8.04506412023109e-05, "loss": 0.0286, "step": 26160 }, { "grad_norm": 0.43449750542640686, "learning_rate": 8.043424526693306e-05, "loss": 0.0335, "step": 26170 }, { "grad_norm": 0.320683091878891, "learning_rate": 8.04178441311751e-05, "loss": 0.0329, "step": 26180 }, { "grad_norm": 0.40727677941322327, "learning_rate": 8.04014377978395e-05, "loss": 0.0305, "step": 26190 }, { "grad_norm": 0.24413223564624786, "learning_rate": 8.038502626972967e-05, "loss": 0.0324, "step": 26200 }, { "grad_norm": 0.41717979311943054, "learning_rate": 8.036860954964989e-05, "loss": 0.0356, "step": 26210 }, { "grad_norm": 0.3159756362438202, "learning_rate": 8.035218764040531e-05, "loss": 0.0371, "step": 26220 }, { "grad_norm": 0.35980597138404846, "learning_rate": 8.033576054480199e-05, "loss": 0.0374, "step": 26230 }, { "grad_norm": 0.44277718663215637, "learning_rate": 8.031932826564688e-05, "loss": 0.028, "step": 26240 }, { "grad_norm": 0.3015069365501404, "learning_rate": 8.030289080574782e-05, "loss": 0.0324, "step": 26250 }, { "grad_norm": 0.46429765224456787, "learning_rate": 8.028644816791349e-05, "loss": 0.0396, "step": 26260 }, { "grad_norm": 0.5446714162826538, "learning_rate": 8.027000035495351e-05, "loss": 0.029, "step": 26270 }, { "grad_norm": 0.4465544819831848, "learning_rate": 8.025354736967836e-05, "loss": 0.0519, "step": 26280 }, { "grad_norm": 0.6059471964836121, "learning_rate": 8.023708921489941e-05, "loss": 0.052, "step": 26290 }, { "grad_norm": 0.4266987442970276, "learning_rate": 8.022062589342887e-05, "loss": 0.0423, "step": 26300 }, { "grad_norm": 0.42609918117523193, "learning_rate": 8.020415740807993e-05, "loss": 0.0441, "step": 26310 }, { "grad_norm": 0.42793309688568115, "learning_rate": 8.018768376166656e-05, "loss": 0.0297, "step": 26320 }, { "grad_norm": 0.32842379808425903, "learning_rate": 8.017120495700368e-05, "loss": 0.0305, "step": 26330 }, { "grad_norm": 0.384535014629364, "learning_rate": 8.015472099690704e-05, "loss": 0.031, "step": 26340 }, { "grad_norm": 0.25122153759002686, "learning_rate": 8.013823188419332e-05, "loss": 0.0299, "step": 26350 }, { "grad_norm": 0.21062734723091125, "learning_rate": 8.012173762168006e-05, "loss": 0.0297, "step": 26360 }, { "grad_norm": 0.34568238258361816, "learning_rate": 8.010523821218567e-05, "loss": 0.0307, "step": 26370 }, { "grad_norm": 0.34408244490623474, "learning_rate": 8.008873365852945e-05, "loss": 0.0303, "step": 26380 }, { "grad_norm": 0.4419579803943634, "learning_rate": 8.007222396353157e-05, "loss": 0.0388, "step": 26390 }, { "grad_norm": 0.39011138677597046, "learning_rate": 8.00557091300131e-05, "loss": 0.0306, "step": 26400 }, { "grad_norm": 0.33489736914634705, "learning_rate": 8.003918916079597e-05, "loss": 0.034, "step": 26410 }, { "grad_norm": 0.28410694003105164, "learning_rate": 8.002266405870298e-05, "loss": 0.0295, "step": 26420 }, { "grad_norm": 0.3582226634025574, "learning_rate": 8.000613382655782e-05, "loss": 0.0431, "step": 26430 }, { "grad_norm": 0.3732755184173584, "learning_rate": 7.998959846718505e-05, "loss": 0.039, "step": 26440 }, { "grad_norm": 0.3266611397266388, "learning_rate": 7.997305798341012e-05, "loss": 0.038, "step": 26450 }, { "grad_norm": 0.34077781438827515, "learning_rate": 7.995651237805937e-05, "loss": 0.0382, "step": 26460 }, { "grad_norm": 0.3813117444515228, "learning_rate": 7.993996165395996e-05, "loss": 0.0354, "step": 26470 }, { "grad_norm": 0.30701905488967896, "learning_rate": 7.992340581393996e-05, "loss": 0.0271, "step": 26480 }, { "grad_norm": 0.4546794295310974, "learning_rate": 7.990684486082831e-05, "loss": 0.0408, "step": 26490 }, { "grad_norm": 0.41458800435066223, "learning_rate": 7.989027879745482e-05, "loss": 0.0362, "step": 26500 }, { "grad_norm": 0.30658158659935, "learning_rate": 7.98737076266502e-05, "loss": 0.0439, "step": 26510 }, { "grad_norm": 0.3757425546646118, "learning_rate": 7.985713135124598e-05, "loss": 0.0357, "step": 26520 }, { "grad_norm": 0.3701249659061432, "learning_rate": 7.98405499740746e-05, "loss": 0.0343, "step": 26530 }, { "grad_norm": 0.3871433138847351, "learning_rate": 7.98239634979694e-05, "loss": 0.0352, "step": 26540 }, { "grad_norm": 0.3422669470310211, "learning_rate": 7.98073719257645e-05, "loss": 0.0291, "step": 26550 }, { "grad_norm": 0.3499075472354889, "learning_rate": 7.979077526029499e-05, "loss": 0.034, "step": 26560 }, { "grad_norm": 0.30622872710227966, "learning_rate": 7.977417350439675e-05, "loss": 0.0274, "step": 26570 }, { "grad_norm": 0.3408692181110382, "learning_rate": 7.97575666609066e-05, "loss": 0.0321, "step": 26580 }, { "grad_norm": 0.36039164662361145, "learning_rate": 7.974095473266216e-05, "loss": 0.0311, "step": 26590 }, { "grad_norm": 0.3556400537490845, "learning_rate": 7.972433772250198e-05, "loss": 0.0338, "step": 26600 }, { "grad_norm": 0.37975507974624634, "learning_rate": 7.970771563326544e-05, "loss": 0.0312, "step": 26610 }, { "grad_norm": 0.43984097242355347, "learning_rate": 7.96910884677928e-05, "loss": 0.0362, "step": 26620 }, { "grad_norm": 0.5190010070800781, "learning_rate": 7.967445622892523e-05, "loss": 0.0384, "step": 26630 }, { "grad_norm": 0.3790876269340515, "learning_rate": 7.965781891950465e-05, "loss": 0.0347, "step": 26640 }, { "grad_norm": 0.5315985083580017, "learning_rate": 7.964117654237397e-05, "loss": 0.032, "step": 26650 }, { "grad_norm": 0.8017279505729675, "learning_rate": 7.962452910037692e-05, "loss": 0.0377, "step": 26660 }, { "grad_norm": 0.33267122507095337, "learning_rate": 7.96078765963581e-05, "loss": 0.0341, "step": 26670 }, { "grad_norm": 0.42216116189956665, "learning_rate": 7.95912190331629e-05, "loss": 0.0448, "step": 26680 }, { "grad_norm": 0.48469626903533936, "learning_rate": 7.957455641363772e-05, "loss": 0.0385, "step": 26690 }, { "grad_norm": 0.45789095759391785, "learning_rate": 7.955788874062968e-05, "loss": 0.0469, "step": 26700 }, { "grad_norm": 0.40269920229911804, "learning_rate": 7.95412160169869e-05, "loss": 0.0343, "step": 26710 }, { "grad_norm": 0.49263808131217957, "learning_rate": 7.952453824555824e-05, "loss": 0.0357, "step": 26720 }, { "grad_norm": 0.5766122937202454, "learning_rate": 7.95078554291935e-05, "loss": 0.0326, "step": 26730 }, { "grad_norm": 0.2853255867958069, "learning_rate": 7.94911675707433e-05, "loss": 0.0292, "step": 26740 }, { "grad_norm": 0.4045121967792511, "learning_rate": 7.947447467305915e-05, "loss": 0.0403, "step": 26750 }, { "grad_norm": 0.3756968379020691, "learning_rate": 7.94577767389934e-05, "loss": 0.0327, "step": 26760 }, { "grad_norm": 0.4070683717727661, "learning_rate": 7.944107377139928e-05, "loss": 0.0259, "step": 26770 }, { "grad_norm": 0.3908686935901642, "learning_rate": 7.942436577313088e-05, "loss": 0.0309, "step": 26780 }, { "grad_norm": 0.30034342408180237, "learning_rate": 7.940765274704312e-05, "loss": 0.0328, "step": 26790 }, { "grad_norm": 0.38086968660354614, "learning_rate": 7.939093469599181e-05, "loss": 0.0309, "step": 26800 }, { "grad_norm": 0.36330264806747437, "learning_rate": 7.93742116228336e-05, "loss": 0.0352, "step": 26810 }, { "grad_norm": 0.27499353885650635, "learning_rate": 7.935748353042602e-05, "loss": 0.0267, "step": 26820 }, { "grad_norm": 0.3530083894729614, "learning_rate": 7.934075042162744e-05, "loss": 0.0357, "step": 26830 }, { "grad_norm": 0.6155182719230652, "learning_rate": 7.932401229929705e-05, "loss": 0.0429, "step": 26840 }, { "grad_norm": 0.5166637301445007, "learning_rate": 7.9307269166295e-05, "loss": 0.0361, "step": 26850 }, { "grad_norm": 0.3788970708847046, "learning_rate": 7.92905210254822e-05, "loss": 0.0332, "step": 26860 }, { "grad_norm": 0.4213002026081085, "learning_rate": 7.927376787972045e-05, "loss": 0.0289, "step": 26870 }, { "grad_norm": 0.3738985061645508, "learning_rate": 7.92570097318724e-05, "loss": 0.0391, "step": 26880 }, { "grad_norm": 0.3528522849082947, "learning_rate": 7.924024658480158e-05, "loss": 0.0349, "step": 26890 }, { "grad_norm": 0.3843112587928772, "learning_rate": 7.922347844137233e-05, "loss": 0.0304, "step": 26900 }, { "grad_norm": 0.350042462348938, "learning_rate": 7.92067053044499e-05, "loss": 0.0321, "step": 26910 }, { "grad_norm": 0.5682858824729919, "learning_rate": 7.918992717690031e-05, "loss": 0.038, "step": 26920 }, { "grad_norm": 0.3693017065525055, "learning_rate": 7.917314406159053e-05, "loss": 0.0337, "step": 26930 }, { "grad_norm": 0.5642910599708557, "learning_rate": 7.915635596138832e-05, "loss": 0.0497, "step": 26940 }, { "grad_norm": 0.40616220235824585, "learning_rate": 7.913956287916228e-05, "loss": 0.0349, "step": 26950 }, { "grad_norm": 0.3483419716358185, "learning_rate": 7.912276481778193e-05, "loss": 0.0256, "step": 26960 }, { "grad_norm": 0.529608964920044, "learning_rate": 7.910596178011759e-05, "loss": 0.0422, "step": 26970 }, { "grad_norm": 0.3143528699874878, "learning_rate": 7.908915376904043e-05, "loss": 0.0248, "step": 26980 }, { "grad_norm": 0.27570104598999023, "learning_rate": 7.907234078742247e-05, "loss": 0.0345, "step": 26990 }, { "grad_norm": 0.2986621856689453, "learning_rate": 7.90555228381366e-05, "loss": 0.0306, "step": 27000 }, { "grad_norm": 0.34111344814300537, "learning_rate": 7.903869992405656e-05, "loss": 0.0212, "step": 27010 }, { "grad_norm": 0.36662840843200684, "learning_rate": 7.902187204805691e-05, "loss": 0.0299, "step": 27020 }, { "grad_norm": 0.33553630113601685, "learning_rate": 7.900503921301308e-05, "loss": 0.0238, "step": 27030 }, { "grad_norm": 0.3213670253753662, "learning_rate": 7.898820142180133e-05, "loss": 0.0397, "step": 27040 }, { "grad_norm": 0.34880709648132324, "learning_rate": 7.897135867729879e-05, "loss": 0.0323, "step": 27050 }, { "grad_norm": 0.39781635999679565, "learning_rate": 7.89545109823834e-05, "loss": 0.0311, "step": 27060 }, { "grad_norm": 0.3649322986602783, "learning_rate": 7.8937658339934e-05, "loss": 0.0312, "step": 27070 }, { "grad_norm": 0.3632924258708954, "learning_rate": 7.892080075283026e-05, "loss": 0.0364, "step": 27080 }, { "grad_norm": 0.4384535253047943, "learning_rate": 7.890393822395263e-05, "loss": 0.0312, "step": 27090 }, { "grad_norm": 0.303181529045105, "learning_rate": 7.88870707561825e-05, "loss": 0.0282, "step": 27100 }, { "grad_norm": 0.6577438712120056, "learning_rate": 7.887019835240203e-05, "loss": 0.0348, "step": 27110 }, { "grad_norm": 0.26681584119796753, "learning_rate": 7.885332101549427e-05, "loss": 0.0333, "step": 27120 }, { "grad_norm": 0.30725356936454773, "learning_rate": 7.883643874834308e-05, "loss": 0.0277, "step": 27130 }, { "grad_norm": 0.3646585941314697, "learning_rate": 7.881955155383321e-05, "loss": 0.0442, "step": 27140 }, { "grad_norm": 0.3719176948070526, "learning_rate": 7.880265943485017e-05, "loss": 0.035, "step": 27150 }, { "grad_norm": 0.4047190845012665, "learning_rate": 7.878576239428038e-05, "loss": 0.0343, "step": 27160 }, { "grad_norm": 0.30828869342803955, "learning_rate": 7.87688604350111e-05, "loss": 0.0376, "step": 27170 }, { "grad_norm": 0.6231196522712708, "learning_rate": 7.875195355993042e-05, "loss": 0.0412, "step": 27180 }, { "grad_norm": 0.3568715751171112, "learning_rate": 7.873504177192724e-05, "loss": 0.0367, "step": 27190 }, { "grad_norm": 0.45607322454452515, "learning_rate": 7.87181250738913e-05, "loss": 0.0456, "step": 27200 }, { "grad_norm": 0.4150049686431885, "learning_rate": 7.870120346871324e-05, "loss": 0.0331, "step": 27210 }, { "grad_norm": 0.40834954380989075, "learning_rate": 7.86842769592845e-05, "loss": 0.0346, "step": 27220 }, { "grad_norm": 0.3929305970668793, "learning_rate": 7.866734554849732e-05, "loss": 0.0304, "step": 27230 }, { "grad_norm": 0.3608088791370392, "learning_rate": 7.865040923924486e-05, "loss": 0.0401, "step": 27240 }, { "grad_norm": 0.3705475628376007, "learning_rate": 7.863346803442104e-05, "loss": 0.0441, "step": 27250 }, { "grad_norm": 0.3198985755443573, "learning_rate": 7.861652193692067e-05, "loss": 0.0294, "step": 27260 }, { "grad_norm": 0.3222581446170807, "learning_rate": 7.859957094963937e-05, "loss": 0.0283, "step": 27270 }, { "grad_norm": 0.3743617832660675, "learning_rate": 7.858261507547357e-05, "loss": 0.0266, "step": 27280 }, { "grad_norm": 0.3467046916484833, "learning_rate": 7.856565431732061e-05, "loss": 0.027, "step": 27290 }, { "grad_norm": 0.34448036551475525, "learning_rate": 7.854868867807859e-05, "loss": 0.0296, "step": 27300 }, { "grad_norm": 0.46931886672973633, "learning_rate": 7.85317181606465e-05, "loss": 0.0284, "step": 27310 }, { "grad_norm": 0.4206221401691437, "learning_rate": 7.85147427679241e-05, "loss": 0.0314, "step": 27320 }, { "grad_norm": 0.35842767357826233, "learning_rate": 7.849776250281205e-05, "loss": 0.0278, "step": 27330 }, { "grad_norm": 0.22190172970294952, "learning_rate": 7.84807773682118e-05, "loss": 0.0312, "step": 27340 }, { "grad_norm": 0.3652127981185913, "learning_rate": 7.846378736702565e-05, "loss": 0.0381, "step": 27350 }, { "grad_norm": 0.36754435300827026, "learning_rate": 7.844679250215671e-05, "loss": 0.0388, "step": 27360 }, { "grad_norm": 0.42570480704307556, "learning_rate": 7.842979277650898e-05, "loss": 0.032, "step": 27370 }, { "grad_norm": 0.30118486285209656, "learning_rate": 7.84127881929872e-05, "loss": 0.0342, "step": 27380 }, { "grad_norm": 0.38835349678993225, "learning_rate": 7.839577875449704e-05, "loss": 0.0283, "step": 27390 }, { "grad_norm": 0.3351316452026367, "learning_rate": 7.837876446394489e-05, "loss": 0.029, "step": 27400 }, { "grad_norm": 0.31969717144966125, "learning_rate": 7.836174532423805e-05, "loss": 0.0362, "step": 27410 }, { "grad_norm": 0.27869969606399536, "learning_rate": 7.834472133828466e-05, "loss": 0.0318, "step": 27420 }, { "grad_norm": 0.4848071038722992, "learning_rate": 7.832769250899359e-05, "loss": 0.0405, "step": 27430 }, { "grad_norm": 0.4935939908027649, "learning_rate": 7.831065883927464e-05, "loss": 0.039, "step": 27440 }, { "grad_norm": 0.397359162569046, "learning_rate": 7.829362033203841e-05, "loss": 0.036, "step": 27450 }, { "grad_norm": 0.27614474296569824, "learning_rate": 7.827657699019628e-05, "loss": 0.0275, "step": 27460 }, { "grad_norm": 0.44236791133880615, "learning_rate": 7.825952881666052e-05, "loss": 0.0463, "step": 27470 }, { "grad_norm": 0.38340964913368225, "learning_rate": 7.824247581434418e-05, "loss": 0.0336, "step": 27480 }, { "grad_norm": 0.4285147190093994, "learning_rate": 7.822541798616116e-05, "loss": 0.034, "step": 27490 }, { "grad_norm": 0.3299403488636017, "learning_rate": 7.820835533502617e-05, "loss": 0.044, "step": 27500 }, { "grad_norm": 0.5219413042068481, "learning_rate": 7.819128786385475e-05, "loss": 0.0372, "step": 27510 }, { "grad_norm": 0.28946104645729065, "learning_rate": 7.817421557556329e-05, "loss": 0.0377, "step": 27520 }, { "grad_norm": 0.35007354617118835, "learning_rate": 7.815713847306893e-05, "loss": 0.0348, "step": 27530 }, { "grad_norm": 0.3846776485443115, "learning_rate": 7.81400565592897e-05, "loss": 0.0308, "step": 27540 }, { "grad_norm": 0.4397721588611603, "learning_rate": 7.812296983714444e-05, "loss": 0.0411, "step": 27550 }, { "grad_norm": 0.41550034284591675, "learning_rate": 7.810587830955281e-05, "loss": 0.035, "step": 27560 }, { "grad_norm": 0.5083597898483276, "learning_rate": 7.808878197943528e-05, "loss": 0.0354, "step": 27570 }, { "grad_norm": 0.4616015553474426, "learning_rate": 7.807168084971312e-05, "loss": 0.0317, "step": 27580 }, { "grad_norm": 0.2872181236743927, "learning_rate": 7.805457492330849e-05, "loss": 0.0299, "step": 27590 }, { "grad_norm": 0.397617906332016, "learning_rate": 7.803746420314428e-05, "loss": 0.0375, "step": 27600 }, { "grad_norm": 0.4354088306427002, "learning_rate": 7.802034869214428e-05, "loss": 0.038, "step": 27610 }, { "grad_norm": 0.44707685708999634, "learning_rate": 7.800322839323303e-05, "loss": 0.0338, "step": 27620 }, { "grad_norm": 0.2727254629135132, "learning_rate": 7.798610330933593e-05, "loss": 0.0277, "step": 27630 }, { "grad_norm": 0.36422038078308105, "learning_rate": 7.796897344337922e-05, "loss": 0.03, "step": 27640 }, { "grad_norm": 0.3403792083263397, "learning_rate": 7.795183879828989e-05, "loss": 0.0363, "step": 27650 }, { "grad_norm": 0.428781658411026, "learning_rate": 7.793469937699579e-05, "loss": 0.0464, "step": 27660 }, { "grad_norm": 0.279598206281662, "learning_rate": 7.791755518242558e-05, "loss": 0.0326, "step": 27670 }, { "grad_norm": 0.45090246200561523, "learning_rate": 7.790040621750876e-05, "loss": 0.0437, "step": 27680 }, { "grad_norm": 0.41827550530433655, "learning_rate": 7.788325248517558e-05, "loss": 0.0296, "step": 27690 }, { "grad_norm": 0.5127187371253967, "learning_rate": 7.786609398835715e-05, "loss": 0.0297, "step": 27700 }, { "grad_norm": 0.29482266306877136, "learning_rate": 7.784893072998541e-05, "loss": 0.0271, "step": 27710 }, { "grad_norm": 0.35666248202323914, "learning_rate": 7.783176271299306e-05, "loss": 0.032, "step": 27720 }, { "grad_norm": 0.32556644082069397, "learning_rate": 7.781458994031368e-05, "loss": 0.0249, "step": 27730 }, { "grad_norm": 0.3495331406593323, "learning_rate": 7.779741241488161e-05, "loss": 0.0275, "step": 27740 }, { "grad_norm": 0.3438538908958435, "learning_rate": 7.7780230139632e-05, "loss": 0.0279, "step": 27750 }, { "grad_norm": 0.39046671986579895, "learning_rate": 7.776304311750087e-05, "loss": 0.0342, "step": 27760 }, { "grad_norm": 0.421658456325531, "learning_rate": 7.7745851351425e-05, "loss": 0.0451, "step": 27770 }, { "grad_norm": 0.3327898383140564, "learning_rate": 7.772865484434197e-05, "loss": 0.0314, "step": 27780 }, { "grad_norm": 0.4081759750843048, "learning_rate": 7.77114535991902e-05, "loss": 0.0464, "step": 27790 }, { "grad_norm": 0.35963326692581177, "learning_rate": 7.769424761890893e-05, "loss": 0.037, "step": 27800 }, { "grad_norm": 0.4388367831707001, "learning_rate": 7.767703690643817e-05, "loss": 0.0387, "step": 27810 }, { "grad_norm": 0.25105276703834534, "learning_rate": 7.76598214647188e-05, "loss": 0.0318, "step": 27820 }, { "grad_norm": 0.43227359652519226, "learning_rate": 7.764260129669241e-05, "loss": 0.0405, "step": 27830 }, { "grad_norm": 0.2714693248271942, "learning_rate": 7.76253764053015e-05, "loss": 0.0405, "step": 27840 }, { "grad_norm": 0.5568122863769531, "learning_rate": 7.760814679348932e-05, "loss": 0.0315, "step": 27850 }, { "grad_norm": 0.2930716276168823, "learning_rate": 7.759091246419992e-05, "loss": 0.0293, "step": 27860 }, { "grad_norm": 0.30492573976516724, "learning_rate": 7.75736734203782e-05, "loss": 0.0258, "step": 27870 }, { "grad_norm": 0.43367746472358704, "learning_rate": 7.755642966496985e-05, "loss": 0.0449, "step": 27880 }, { "grad_norm": 0.36225730180740356, "learning_rate": 7.753918120092132e-05, "loss": 0.0369, "step": 27890 }, { "grad_norm": 0.46163544058799744, "learning_rate": 7.752192803117993e-05, "loss": 0.0322, "step": 27900 }, { "grad_norm": 0.43511682748794556, "learning_rate": 7.750467015869377e-05, "loss": 0.0338, "step": 27910 }, { "grad_norm": 0.2902391850948334, "learning_rate": 7.748740758641174e-05, "loss": 0.0349, "step": 27920 }, { "grad_norm": 0.299284428358078, "learning_rate": 7.74701403172835e-05, "loss": 0.036, "step": 27930 }, { "grad_norm": 0.3735094368457794, "learning_rate": 7.745286835425962e-05, "loss": 0.0374, "step": 27940 }, { "grad_norm": 0.45697230100631714, "learning_rate": 7.743559170029138e-05, "loss": 0.0332, "step": 27950 }, { "grad_norm": 0.36422446370124817, "learning_rate": 7.741831035833087e-05, "loss": 0.0313, "step": 27960 }, { "grad_norm": 0.28848475217819214, "learning_rate": 7.740102433133102e-05, "loss": 0.0275, "step": 27970 }, { "grad_norm": 0.36703261733055115, "learning_rate": 7.738373362224553e-05, "loss": 0.0273, "step": 27980 }, { "grad_norm": 0.359262079000473, "learning_rate": 7.73664382340289e-05, "loss": 0.032, "step": 27990 }, { "grad_norm": 0.38975459337234497, "learning_rate": 7.734913816963647e-05, "loss": 0.0364, "step": 28000 }, { "grad_norm": 0.24892401695251465, "learning_rate": 7.73318334320243e-05, "loss": 0.026, "step": 28010 }, { "grad_norm": 0.44241881370544434, "learning_rate": 7.731452402414934e-05, "loss": 0.0378, "step": 28020 }, { "grad_norm": 0.4700292944908142, "learning_rate": 7.729720994896928e-05, "loss": 0.0292, "step": 28030 }, { "grad_norm": 0.4952000677585602, "learning_rate": 7.727989120944262e-05, "loss": 0.0316, "step": 28040 }, { "grad_norm": 0.3716523051261902, "learning_rate": 7.726256780852865e-05, "loss": 0.0291, "step": 28050 }, { "grad_norm": 0.4033032953739166, "learning_rate": 7.724523974918749e-05, "loss": 0.0292, "step": 28060 }, { "grad_norm": 0.48343929648399353, "learning_rate": 7.722790703438002e-05, "loss": 0.0289, "step": 28070 }, { "grad_norm": 0.5142935514450073, "learning_rate": 7.72105696670679e-05, "loss": 0.0266, "step": 28080 }, { "grad_norm": 0.42130592465400696, "learning_rate": 7.719322765021364e-05, "loss": 0.0324, "step": 28090 }, { "grad_norm": 0.32407379150390625, "learning_rate": 7.717588098678051e-05, "loss": 0.0299, "step": 28100 }, { "grad_norm": 0.3545677959918976, "learning_rate": 7.715852967973258e-05, "loss": 0.0292, "step": 28110 }, { "grad_norm": 0.35939663648605347, "learning_rate": 7.714117373203474e-05, "loss": 0.0318, "step": 28120 }, { "grad_norm": 0.593889594078064, "learning_rate": 7.712381314665259e-05, "loss": 0.0296, "step": 28130 }, { "grad_norm": 0.3044724762439728, "learning_rate": 7.710644792655261e-05, "loss": 0.0282, "step": 28140 }, { "grad_norm": 0.5878618359565735, "learning_rate": 7.708907807470207e-05, "loss": 0.0458, "step": 28150 }, { "grad_norm": 0.4029681980609894, "learning_rate": 7.707170359406896e-05, "loss": 0.035, "step": 28160 }, { "grad_norm": 0.4284931719303131, "learning_rate": 7.705432448762213e-05, "loss": 0.0289, "step": 28170 }, { "grad_norm": 0.3611915409564972, "learning_rate": 7.703694075833117e-05, "loss": 0.0442, "step": 28180 }, { "grad_norm": 0.3248620629310608, "learning_rate": 7.70195524091665e-05, "loss": 0.0426, "step": 28190 }, { "grad_norm": 0.41004252433776855, "learning_rate": 7.70021594430993e-05, "loss": 0.038, "step": 28200 }, { "grad_norm": 0.3743242919445038, "learning_rate": 7.698476186310157e-05, "loss": 0.0259, "step": 28210 }, { "grad_norm": 0.31225284934043884, "learning_rate": 7.696735967214608e-05, "loss": 0.031, "step": 28220 }, { "grad_norm": 0.32438018918037415, "learning_rate": 7.694995287320636e-05, "loss": 0.0386, "step": 28230 }, { "grad_norm": 0.404082328081131, "learning_rate": 7.693254146925679e-05, "loss": 0.0316, "step": 28240 }, { "grad_norm": 0.4079211354255676, "learning_rate": 7.691512546327251e-05, "loss": 0.0398, "step": 28250 }, { "grad_norm": 0.371519535779953, "learning_rate": 7.689770485822939e-05, "loss": 0.0367, "step": 28260 }, { "grad_norm": 0.369473934173584, "learning_rate": 7.688027965710416e-05, "loss": 0.0328, "step": 28270 }, { "grad_norm": 0.2856542468070984, "learning_rate": 7.686284986287433e-05, "loss": 0.0344, "step": 28280 }, { "grad_norm": 0.31816357374191284, "learning_rate": 7.684541547851817e-05, "loss": 0.0294, "step": 28290 }, { "grad_norm": 0.21600249409675598, "learning_rate": 7.68279765070147e-05, "loss": 0.032, "step": 28300 }, { "grad_norm": 0.45750948786735535, "learning_rate": 7.68105329513438e-05, "loss": 0.0292, "step": 28310 }, { "grad_norm": 0.3270406424999237, "learning_rate": 7.67930848144861e-05, "loss": 0.0294, "step": 28320 }, { "grad_norm": 0.33716583251953125, "learning_rate": 7.6775632099423e-05, "loss": 0.0319, "step": 28330 }, { "grad_norm": 0.3676215410232544, "learning_rate": 7.675817480913667e-05, "loss": 0.0387, "step": 28340 }, { "grad_norm": 0.3120959997177124, "learning_rate": 7.674071294661011e-05, "loss": 0.0235, "step": 28350 }, { "grad_norm": 0.4065955877304077, "learning_rate": 7.672324651482707e-05, "loss": 0.027, "step": 28360 }, { "grad_norm": 0.3052958846092224, "learning_rate": 7.670577551677209e-05, "loss": 0.0315, "step": 28370 }, { "grad_norm": 0.4089806079864502, "learning_rate": 7.668829995543047e-05, "loss": 0.0292, "step": 28380 }, { "grad_norm": 0.2404148429632187, "learning_rate": 7.667081983378832e-05, "loss": 0.0324, "step": 28390 }, { "grad_norm": 1.0936790704727173, "learning_rate": 7.66533351548325e-05, "loss": 0.026, "step": 28400 }, { "grad_norm": 0.4844784438610077, "learning_rate": 7.663584592155069e-05, "loss": 0.0355, "step": 28410 }, { "grad_norm": 0.38630566000938416, "learning_rate": 7.661835213693129e-05, "loss": 0.0369, "step": 28420 }, { "grad_norm": 0.4219539761543274, "learning_rate": 7.660085380396353e-05, "loss": 0.033, "step": 28430 }, { "grad_norm": 0.5011857151985168, "learning_rate": 7.658335092563738e-05, "loss": 0.0303, "step": 28440 }, { "grad_norm": 0.23154199123382568, "learning_rate": 7.656584350494362e-05, "loss": 0.0289, "step": 28450 }, { "grad_norm": 0.40044888854026794, "learning_rate": 7.654833154487378e-05, "loss": 0.033, "step": 28460 }, { "grad_norm": 0.3691924810409546, "learning_rate": 7.653081504842017e-05, "loss": 0.0319, "step": 28470 }, { "grad_norm": 0.44115373492240906, "learning_rate": 7.65132940185759e-05, "loss": 0.0385, "step": 28480 }, { "grad_norm": 0.3983715772628784, "learning_rate": 7.649576845833481e-05, "loss": 0.0298, "step": 28490 }, { "grad_norm": 0.4103313684463501, "learning_rate": 7.647823837069156e-05, "loss": 0.0366, "step": 28500 }, { "grad_norm": 0.34479814767837524, "learning_rate": 7.646070375864156e-05, "loss": 0.0327, "step": 28510 }, { "grad_norm": 0.5205528140068054, "learning_rate": 7.644316462518099e-05, "loss": 0.0391, "step": 28520 }, { "grad_norm": 0.37007713317871094, "learning_rate": 7.642562097330679e-05, "loss": 0.0307, "step": 28530 }, { "grad_norm": 0.24161161482334137, "learning_rate": 7.640807280601671e-05, "loss": 0.0328, "step": 28540 }, { "grad_norm": 0.3945372700691223, "learning_rate": 7.639052012630927e-05, "loss": 0.0421, "step": 28550 }, { "grad_norm": 0.39486581087112427, "learning_rate": 7.63729629371837e-05, "loss": 0.0344, "step": 28560 }, { "grad_norm": 0.35408157110214233, "learning_rate": 7.635540124164009e-05, "loss": 0.0435, "step": 28570 }, { "grad_norm": 0.35301080346107483, "learning_rate": 7.633783504267922e-05, "loss": 0.0406, "step": 28580 }, { "grad_norm": 0.44568952918052673, "learning_rate": 7.632026434330269e-05, "loss": 0.0366, "step": 28590 }, { "grad_norm": 0.27382874488830566, "learning_rate": 7.630268914651282e-05, "loss": 0.035, "step": 28600 }, { "grad_norm": 0.44911089539527893, "learning_rate": 7.628510945531278e-05, "loss": 0.0457, "step": 28610 }, { "grad_norm": 0.3210890591144562, "learning_rate": 7.626752527270641e-05, "loss": 0.0349, "step": 28620 }, { "grad_norm": 0.33935078978538513, "learning_rate": 7.62499366016984e-05, "loss": 0.0273, "step": 28630 }, { "grad_norm": 0.2615303695201874, "learning_rate": 7.623234344529416e-05, "loss": 0.0334, "step": 28640 }, { "grad_norm": 0.20795287191867828, "learning_rate": 7.62147458064999e-05, "loss": 0.0231, "step": 28650 }, { "grad_norm": 0.5482035875320435, "learning_rate": 7.619714368832254e-05, "loss": 0.0344, "step": 28660 }, { "grad_norm": 0.4589528441429138, "learning_rate": 7.61795370937698e-05, "loss": 0.0282, "step": 28670 }, { "grad_norm": 0.275250643491745, "learning_rate": 7.61619260258502e-05, "loss": 0.0324, "step": 28680 }, { "grad_norm": 0.3480728268623352, "learning_rate": 7.614431048757298e-05, "loss": 0.0363, "step": 28690 }, { "grad_norm": 0.24400921165943146, "learning_rate": 7.612669048194814e-05, "loss": 0.0326, "step": 28700 }, { "grad_norm": 0.2978620231151581, "learning_rate": 7.610906601198646e-05, "loss": 0.0291, "step": 28710 }, { "grad_norm": 0.4894084930419922, "learning_rate": 7.60914370806995e-05, "loss": 0.0339, "step": 28720 }, { "grad_norm": 0.37824034690856934, "learning_rate": 7.607380369109953e-05, "loss": 0.0517, "step": 28730 }, { "grad_norm": 0.3368555009365082, "learning_rate": 7.605616584619961e-05, "loss": 0.0294, "step": 28740 }, { "grad_norm": 0.43392080068588257, "learning_rate": 7.603852354901362e-05, "loss": 0.035, "step": 28750 }, { "grad_norm": 0.36914780735969543, "learning_rate": 7.602087680255609e-05, "loss": 0.0458, "step": 28760 }, { "grad_norm": 0.386322945356369, "learning_rate": 7.600322560984238e-05, "loss": 0.0324, "step": 28770 }, { "grad_norm": 0.3379107117652893, "learning_rate": 7.598556997388863e-05, "loss": 0.0323, "step": 28780 }, { "grad_norm": 0.6491225957870483, "learning_rate": 7.596790989771166e-05, "loss": 0.0416, "step": 28790 }, { "grad_norm": 0.28820866346359253, "learning_rate": 7.595024538432914e-05, "loss": 0.0327, "step": 28800 }, { "grad_norm": 0.38758227229118347, "learning_rate": 7.59325764367594e-05, "loss": 0.0288, "step": 28810 }, { "grad_norm": 0.3663051128387451, "learning_rate": 7.59149030580216e-05, "loss": 0.0291, "step": 28820 }, { "grad_norm": 0.3313526511192322, "learning_rate": 7.589722525113562e-05, "loss": 0.027, "step": 28830 }, { "grad_norm": 0.3020070791244507, "learning_rate": 7.587954301912216e-05, "loss": 0.0243, "step": 28840 }, { "grad_norm": 0.5440999269485474, "learning_rate": 7.586185636500263e-05, "loss": 0.0348, "step": 28850 }, { "grad_norm": 0.34001827239990234, "learning_rate": 7.584416529179914e-05, "loss": 0.0366, "step": 28860 }, { "grad_norm": 0.3134031593799591, "learning_rate": 7.582646980253465e-05, "loss": 0.0346, "step": 28870 }, { "grad_norm": 0.415707528591156, "learning_rate": 7.580876990023282e-05, "loss": 0.0412, "step": 28880 }, { "grad_norm": 0.3802056312561035, "learning_rate": 7.579106558791809e-05, "loss": 0.0342, "step": 28890 }, { "grad_norm": 0.2933587431907654, "learning_rate": 7.577335686861565e-05, "loss": 0.0297, "step": 28900 }, { "grad_norm": 0.6040385961532593, "learning_rate": 7.575564374535141e-05, "loss": 0.0329, "step": 28910 }, { "grad_norm": 0.2945995032787323, "learning_rate": 7.573792622115207e-05, "loss": 0.026, "step": 28920 }, { "grad_norm": 0.45280107855796814, "learning_rate": 7.572020429904507e-05, "loss": 0.0299, "step": 28930 }, { "grad_norm": 0.47760671377182007, "learning_rate": 7.570247798205861e-05, "loss": 0.0364, "step": 28940 }, { "grad_norm": 0.3466549515724182, "learning_rate": 7.568474727322164e-05, "loss": 0.0329, "step": 28950 }, { "grad_norm": 0.36253809928894043, "learning_rate": 7.566701217556384e-05, "loss": 0.0272, "step": 28960 }, { "grad_norm": 0.3950912356376648, "learning_rate": 7.564927269211564e-05, "loss": 0.0303, "step": 28970 }, { "grad_norm": 0.3312263488769531, "learning_rate": 7.563152882590824e-05, "loss": 0.0247, "step": 28980 }, { "grad_norm": 0.3734218180179596, "learning_rate": 7.56137805799736e-05, "loss": 0.0315, "step": 28990 }, { "grad_norm": 0.45165470242500305, "learning_rate": 7.559602795734439e-05, "loss": 0.0344, "step": 29000 }, { "grad_norm": 0.3384822607040405, "learning_rate": 7.557827096105408e-05, "loss": 0.0329, "step": 29010 }, { "grad_norm": 0.5073298215866089, "learning_rate": 7.55605095941368e-05, "loss": 0.0374, "step": 29020 }, { "grad_norm": 0.34890216588974, "learning_rate": 7.554274385962752e-05, "loss": 0.0328, "step": 29030 }, { "grad_norm": 0.4253486692905426, "learning_rate": 7.552497376056191e-05, "loss": 0.0319, "step": 29040 }, { "grad_norm": 0.2505412995815277, "learning_rate": 7.550719929997639e-05, "loss": 0.0306, "step": 29050 }, { "grad_norm": 0.412494033575058, "learning_rate": 7.548942048090813e-05, "loss": 0.0293, "step": 29060 }, { "grad_norm": 0.4322075843811035, "learning_rate": 7.547163730639506e-05, "loss": 0.0279, "step": 29070 }, { "grad_norm": 0.5330989360809326, "learning_rate": 7.545384977947583e-05, "loss": 0.0323, "step": 29080 }, { "grad_norm": 0.42833831906318665, "learning_rate": 7.543605790318981e-05, "loss": 0.03, "step": 29090 }, { "grad_norm": 0.25554537773132324, "learning_rate": 7.54182616805772e-05, "loss": 0.0306, "step": 29100 }, { "grad_norm": 0.25401297211647034, "learning_rate": 7.540046111467885e-05, "loss": 0.0316, "step": 29110 }, { "grad_norm": 0.3051159381866455, "learning_rate": 7.53826562085364e-05, "loss": 0.0339, "step": 29120 }, { "grad_norm": 0.37498342990875244, "learning_rate": 7.536484696519221e-05, "loss": 0.021, "step": 29130 }, { "grad_norm": 0.3872300982475281, "learning_rate": 7.534703338768942e-05, "loss": 0.0324, "step": 29140 }, { "grad_norm": 0.2781304717063904, "learning_rate": 7.532921547907185e-05, "loss": 0.0343, "step": 29150 }, { "grad_norm": 0.3586212694644928, "learning_rate": 7.531139324238412e-05, "loss": 0.0434, "step": 29160 }, { "grad_norm": 0.4675503075122833, "learning_rate": 7.529356668067157e-05, "loss": 0.042, "step": 29170 }, { "grad_norm": 0.5243116617202759, "learning_rate": 7.527573579698023e-05, "loss": 0.0362, "step": 29180 }, { "grad_norm": 0.45435044169425964, "learning_rate": 7.525790059435693e-05, "loss": 0.0295, "step": 29190 }, { "grad_norm": 0.38925614953041077, "learning_rate": 7.524006107584926e-05, "loss": 0.0361, "step": 29200 }, { "grad_norm": 0.4698939323425293, "learning_rate": 7.522221724450544e-05, "loss": 0.03, "step": 29210 }, { "grad_norm": 0.28976887464523315, "learning_rate": 7.520436910337451e-05, "loss": 0.0374, "step": 29220 }, { "grad_norm": 0.2995370030403137, "learning_rate": 7.518651665550627e-05, "loss": 0.0272, "step": 29230 }, { "grad_norm": 0.25795358419418335, "learning_rate": 7.516865990395117e-05, "loss": 0.0345, "step": 29240 }, { "grad_norm": 0.3573358654975891, "learning_rate": 7.515079885176047e-05, "loss": 0.0299, "step": 29250 }, { "grad_norm": 0.361053466796875, "learning_rate": 7.513293350198612e-05, "loss": 0.033, "step": 29260 }, { "grad_norm": 0.3322802484035492, "learning_rate": 7.511506385768081e-05, "loss": 0.0364, "step": 29270 }, { "grad_norm": 0.4111112952232361, "learning_rate": 7.509718992189801e-05, "loss": 0.0276, "step": 29280 }, { "grad_norm": 0.46299222111701965, "learning_rate": 7.507931169769182e-05, "loss": 0.0318, "step": 29290 }, { "grad_norm": 0.3642261326313019, "learning_rate": 7.506142918811722e-05, "loss": 0.0305, "step": 29300 }, { "grad_norm": 0.3222137689590454, "learning_rate": 7.504354239622978e-05, "loss": 0.0427, "step": 29310 }, { "grad_norm": 0.45341554284095764, "learning_rate": 7.50256513250859e-05, "loss": 0.0299, "step": 29320 }, { "grad_norm": 0.456906795501709, "learning_rate": 7.500775597774265e-05, "loss": 0.0308, "step": 29330 }, { "grad_norm": 0.32442212104797363, "learning_rate": 7.498985635725788e-05, "loss": 0.0273, "step": 29340 }, { "grad_norm": 0.4490669071674347, "learning_rate": 7.497195246669012e-05, "loss": 0.0384, "step": 29350 }, { "grad_norm": 0.30037471652030945, "learning_rate": 7.495404430909868e-05, "loss": 0.0296, "step": 29360 }, { "grad_norm": 0.4019584357738495, "learning_rate": 7.493613188754356e-05, "loss": 0.0308, "step": 29370 }, { "grad_norm": 0.39285996556282043, "learning_rate": 7.49182152050855e-05, "loss": 0.0283, "step": 29380 }, { "grad_norm": 0.4888645112514496, "learning_rate": 7.490029426478598e-05, "loss": 0.0242, "step": 29390 }, { "grad_norm": 0.31295743584632874, "learning_rate": 7.488236906970719e-05, "loss": 0.0292, "step": 29400 }, { "grad_norm": 0.4868641793727875, "learning_rate": 7.486443962291207e-05, "loss": 0.0307, "step": 29410 }, { "grad_norm": 0.44441038370132446, "learning_rate": 7.484650592746424e-05, "loss": 0.0302, "step": 29420 }, { "grad_norm": 0.3693835437297821, "learning_rate": 7.482856798642811e-05, "loss": 0.0292, "step": 29430 }, { "grad_norm": 0.26993095874786377, "learning_rate": 7.481062580286878e-05, "loss": 0.0287, "step": 29440 }, { "grad_norm": 0.3039804995059967, "learning_rate": 7.479267937985208e-05, "loss": 0.0372, "step": 29450 }, { "grad_norm": 0.42921558022499084, "learning_rate": 7.477472872044456e-05, "loss": 0.0279, "step": 29460 }, { "grad_norm": 0.34591519832611084, "learning_rate": 7.475677382771347e-05, "loss": 0.0321, "step": 29470 }, { "grad_norm": 0.3401302099227905, "learning_rate": 7.473881470472683e-05, "loss": 0.0337, "step": 29480 }, { "grad_norm": 0.417194664478302, "learning_rate": 7.472085135455337e-05, "loss": 0.0357, "step": 29490 }, { "grad_norm": 0.35517576336860657, "learning_rate": 7.470288378026256e-05, "loss": 0.038, "step": 29500 }, { "grad_norm": 0.3729175925254822, "learning_rate": 7.468491198492451e-05, "loss": 0.0319, "step": 29510 }, { "grad_norm": 0.40090373158454895, "learning_rate": 7.466693597161013e-05, "loss": 0.0312, "step": 29520 }, { "grad_norm": 0.3250119686126709, "learning_rate": 7.464895574339104e-05, "loss": 0.0362, "step": 29530 }, { "grad_norm": 0.2766861319541931, "learning_rate": 7.463097130333958e-05, "loss": 0.0408, "step": 29540 }, { "grad_norm": 0.2767329812049866, "learning_rate": 7.461298265452876e-05, "loss": 0.0309, "step": 29550 }, { "grad_norm": 0.3015795052051544, "learning_rate": 7.459498980003239e-05, "loss": 0.0275, "step": 29560 }, { "grad_norm": 0.5304531455039978, "learning_rate": 7.457699274292493e-05, "loss": 0.033, "step": 29570 }, { "grad_norm": 0.28604936599731445, "learning_rate": 7.455899148628159e-05, "loss": 0.035, "step": 29580 }, { "grad_norm": 0.4154633581638336, "learning_rate": 7.45409860331783e-05, "loss": 0.0327, "step": 29590 }, { "grad_norm": 0.5925370454788208, "learning_rate": 7.452297638669169e-05, "loss": 0.0277, "step": 29600 }, { "grad_norm": 0.3642883002758026, "learning_rate": 7.450496254989911e-05, "loss": 0.0206, "step": 29610 }, { "grad_norm": 0.3429102599620819, "learning_rate": 7.448694452587866e-05, "loss": 0.026, "step": 29620 }, { "grad_norm": 0.4094419479370117, "learning_rate": 7.44689223177091e-05, "loss": 0.0344, "step": 29630 }, { "grad_norm": 0.3251003324985504, "learning_rate": 7.445089592846994e-05, "loss": 0.0266, "step": 29640 }, { "grad_norm": 0.32563528418540955, "learning_rate": 7.443286536124141e-05, "loss": 0.0361, "step": 29650 }, { "grad_norm": 0.35364022850990295, "learning_rate": 7.441483061910443e-05, "loss": 0.0327, "step": 29660 }, { "grad_norm": 0.4127466082572937, "learning_rate": 7.439679170514064e-05, "loss": 0.0364, "step": 29670 }, { "grad_norm": 0.3812296986579895, "learning_rate": 7.43787486224324e-05, "loss": 0.0312, "step": 29680 }, { "grad_norm": 0.3551247715950012, "learning_rate": 7.436070137406276e-05, "loss": 0.0301, "step": 29690 }, { "grad_norm": 0.40541988611221313, "learning_rate": 7.434264996311556e-05, "loss": 0.0293, "step": 29700 }, { "grad_norm": 0.5155392289161682, "learning_rate": 7.432459439267525e-05, "loss": 0.0412, "step": 29710 }, { "grad_norm": 0.323551744222641, "learning_rate": 7.430653466582701e-05, "loss": 0.0321, "step": 29720 }, { "grad_norm": 0.35828396677970886, "learning_rate": 7.42884707856568e-05, "loss": 0.04, "step": 29730 }, { "grad_norm": 0.29359474778175354, "learning_rate": 7.427040275525122e-05, "loss": 0.0205, "step": 29740 }, { "grad_norm": 0.25087839365005493, "learning_rate": 7.42523305776976e-05, "loss": 0.0382, "step": 29750 }, { "grad_norm": 0.3467518389225006, "learning_rate": 7.4234254256084e-05, "loss": 0.0277, "step": 29760 }, { "grad_norm": 0.27143627405166626, "learning_rate": 7.421617379349915e-05, "loss": 0.0359, "step": 29770 }, { "grad_norm": 0.3229122757911682, "learning_rate": 7.41980891930325e-05, "loss": 0.0257, "step": 29780 }, { "grad_norm": 0.5434967875480652, "learning_rate": 7.418000045777425e-05, "loss": 0.0333, "step": 29790 }, { "grad_norm": 0.33345478773117065, "learning_rate": 7.416190759081523e-05, "loss": 0.0333, "step": 29800 }, { "grad_norm": 0.4347216784954071, "learning_rate": 7.414381059524704e-05, "loss": 0.0256, "step": 29810 }, { "grad_norm": 0.41394269466400146, "learning_rate": 7.412570947416195e-05, "loss": 0.0319, "step": 29820 }, { "grad_norm": 0.6292665600776672, "learning_rate": 7.410760423065295e-05, "loss": 0.0415, "step": 29830 }, { "grad_norm": 0.6235424280166626, "learning_rate": 7.408949486781372e-05, "loss": 0.0444, "step": 29840 }, { "grad_norm": 0.4899763762950897, "learning_rate": 7.407138138873868e-05, "loss": 0.0394, "step": 29850 }, { "grad_norm": 0.4263523519039154, "learning_rate": 7.405326379652292e-05, "loss": 0.0325, "step": 29860 }, { "grad_norm": 0.45352670550346375, "learning_rate": 7.403514209426222e-05, "loss": 0.0311, "step": 29870 }, { "grad_norm": 0.4981698989868164, "learning_rate": 7.40170162850531e-05, "loss": 0.0336, "step": 29880 }, { "grad_norm": 0.30525922775268555, "learning_rate": 7.399888637199278e-05, "loss": 0.0299, "step": 29890 }, { "grad_norm": 0.4711129069328308, "learning_rate": 7.398075235817914e-05, "loss": 0.0403, "step": 29900 }, { "grad_norm": 0.31089645624160767, "learning_rate": 7.39626142467108e-05, "loss": 0.0278, "step": 29910 }, { "grad_norm": 0.3227521777153015, "learning_rate": 7.394447204068706e-05, "loss": 0.0276, "step": 29920 }, { "grad_norm": 0.41274866461753845, "learning_rate": 7.392632574320793e-05, "loss": 0.0378, "step": 29930 }, { "grad_norm": 0.3012392818927765, "learning_rate": 7.390817535737411e-05, "loss": 0.0274, "step": 29940 }, { "grad_norm": 0.33487942814826965, "learning_rate": 7.389002088628703e-05, "loss": 0.0313, "step": 29950 }, { "grad_norm": 0.35451197624206543, "learning_rate": 7.387186233304877e-05, "loss": 0.0303, "step": 29960 }, { "grad_norm": 0.2901807725429535, "learning_rate": 7.385369970076212e-05, "loss": 0.0342, "step": 29970 }, { "grad_norm": 0.3659496009349823, "learning_rate": 7.38355329925306e-05, "loss": 0.0327, "step": 29980 }, { "grad_norm": 0.3725307583808899, "learning_rate": 7.381736221145838e-05, "loss": 0.0325, "step": 29990 }, { "grad_norm": 0.44051337242126465, "learning_rate": 7.37991873606504e-05, "loss": 0.0251, "step": 30000 }, { "grad_norm": 0.24417923390865326, "learning_rate": 7.378100844321218e-05, "loss": 0.0309, "step": 30010 }, { "grad_norm": 0.4035658836364746, "learning_rate": 7.376282546225004e-05, "loss": 0.0264, "step": 30020 }, { "grad_norm": 0.40479201078414917, "learning_rate": 7.374463842087094e-05, "loss": 0.0285, "step": 30030 }, { "grad_norm": 0.3132748007774353, "learning_rate": 7.372644732218254e-05, "loss": 0.0295, "step": 30040 }, { "grad_norm": 0.3927144408226013, "learning_rate": 7.370825216929322e-05, "loss": 0.0408, "step": 30050 }, { "grad_norm": 0.413693904876709, "learning_rate": 7.369005296531205e-05, "loss": 0.0378, "step": 30060 }, { "grad_norm": 0.32255449891090393, "learning_rate": 7.367184971334873e-05, "loss": 0.031, "step": 30070 }, { "grad_norm": 0.2965903878211975, "learning_rate": 7.365364241651371e-05, "loss": 0.0322, "step": 30080 }, { "grad_norm": 0.31066349148750305, "learning_rate": 7.363543107791815e-05, "loss": 0.026, "step": 30090 }, { "grad_norm": 0.37831154465675354, "learning_rate": 7.361721570067384e-05, "loss": 0.0258, "step": 30100 }, { "grad_norm": 0.32464540004730225, "learning_rate": 7.359899628789331e-05, "loss": 0.0273, "step": 30110 }, { "grad_norm": 0.3968704044818878, "learning_rate": 7.358077284268974e-05, "loss": 0.0338, "step": 30120 }, { "grad_norm": 0.367382287979126, "learning_rate": 7.356254536817702e-05, "loss": 0.0274, "step": 30130 }, { "grad_norm": 0.44282180070877075, "learning_rate": 7.354431386746973e-05, "loss": 0.0249, "step": 30140 }, { "grad_norm": 0.3696334958076477, "learning_rate": 7.352607834368316e-05, "loss": 0.0282, "step": 30150 }, { "grad_norm": 0.49292364716529846, "learning_rate": 7.350783879993324e-05, "loss": 0.0255, "step": 30160 }, { "grad_norm": 0.28867071866989136, "learning_rate": 7.348959523933658e-05, "loss": 0.0317, "step": 30170 }, { "grad_norm": 0.4148137867450714, "learning_rate": 7.347134766501057e-05, "loss": 0.0251, "step": 30180 }, { "grad_norm": 0.35290342569351196, "learning_rate": 7.345309608007315e-05, "loss": 0.0217, "step": 30190 }, { "grad_norm": 0.3043115735054016, "learning_rate": 7.343484048764308e-05, "loss": 0.031, "step": 30200 }, { "grad_norm": 0.3383914828300476, "learning_rate": 7.341658089083972e-05, "loss": 0.0286, "step": 30210 }, { "grad_norm": 0.24671810865402222, "learning_rate": 7.339831729278313e-05, "loss": 0.025, "step": 30220 }, { "grad_norm": 0.2871514558792114, "learning_rate": 7.338004969659404e-05, "loss": 0.0294, "step": 30230 }, { "grad_norm": 0.37312620878219604, "learning_rate": 7.336177810539391e-05, "loss": 0.0309, "step": 30240 }, { "grad_norm": 0.337047815322876, "learning_rate": 7.334350252230485e-05, "loss": 0.0255, "step": 30250 }, { "grad_norm": 0.24208037555217743, "learning_rate": 7.332522295044965e-05, "loss": 0.0235, "step": 30260 }, { "grad_norm": 0.44625982642173767, "learning_rate": 7.33069393929518e-05, "loss": 0.0281, "step": 30270 }, { "grad_norm": 0.43273454904556274, "learning_rate": 7.328865185293545e-05, "loss": 0.0427, "step": 30280 }, { "grad_norm": 0.28942444920539856, "learning_rate": 7.327036033352546e-05, "loss": 0.0325, "step": 30290 }, { "grad_norm": 0.4291757047176361, "learning_rate": 7.325206483784733e-05, "loss": 0.0306, "step": 30300 }, { "grad_norm": 0.38570505380630493, "learning_rate": 7.323376536902724e-05, "loss": 0.0384, "step": 30310 }, { "grad_norm": 0.855404257774353, "learning_rate": 7.321546193019213e-05, "loss": 0.0404, "step": 30320 }, { "grad_norm": 0.34707340598106384, "learning_rate": 7.31971545244695e-05, "loss": 0.0434, "step": 30330 }, { "grad_norm": 0.30068719387054443, "learning_rate": 7.31788431549876e-05, "loss": 0.0336, "step": 30340 }, { "grad_norm": 0.5328952074050903, "learning_rate": 7.316052782487534e-05, "loss": 0.0282, "step": 30350 }, { "grad_norm": 0.27400121092796326, "learning_rate": 7.314220853726234e-05, "loss": 0.0328, "step": 30360 }, { "grad_norm": 0.3913022577762604, "learning_rate": 7.312388529527884e-05, "loss": 0.0341, "step": 30370 }, { "grad_norm": 0.3502879738807678, "learning_rate": 7.310555810205577e-05, "loss": 0.0345, "step": 30380 }, { "grad_norm": 0.2750985026359558, "learning_rate": 7.308722696072476e-05, "loss": 0.031, "step": 30390 }, { "grad_norm": 0.30412641167640686, "learning_rate": 7.306889187441811e-05, "loss": 0.0254, "step": 30400 }, { "grad_norm": 0.20522330701351166, "learning_rate": 7.305055284626876e-05, "loss": 0.0247, "step": 30410 }, { "grad_norm": 0.38876813650131226, "learning_rate": 7.303220987941037e-05, "loss": 0.031, "step": 30420 }, { "grad_norm": 0.26496267318725586, "learning_rate": 7.301386297697726e-05, "loss": 0.0323, "step": 30430 }, { "grad_norm": 0.35590052604675293, "learning_rate": 7.299551214210438e-05, "loss": 0.0312, "step": 30440 }, { "grad_norm": 0.3085464835166931, "learning_rate": 7.297715737792738e-05, "loss": 0.0284, "step": 30450 }, { "grad_norm": 0.2566228210926056, "learning_rate": 7.295879868758265e-05, "loss": 0.0316, "step": 30460 }, { "grad_norm": 0.374467134475708, "learning_rate": 7.294043607420713e-05, "loss": 0.0333, "step": 30470 }, { "grad_norm": 0.3548474907875061, "learning_rate": 7.292206954093852e-05, "loss": 0.0273, "step": 30480 }, { "grad_norm": 0.40145111083984375, "learning_rate": 7.290369909091515e-05, "loss": 0.0274, "step": 30490 }, { "grad_norm": 0.3980253338813782, "learning_rate": 7.2885324727276e-05, "loss": 0.0322, "step": 30500 }, { "grad_norm": 0.39706161618232727, "learning_rate": 7.286694645316076e-05, "loss": 0.0313, "step": 30510 }, { "grad_norm": 0.47387707233428955, "learning_rate": 7.284856427170982e-05, "loss": 0.0279, "step": 30520 }, { "grad_norm": 0.36592841148376465, "learning_rate": 7.283017818606414e-05, "loss": 0.037, "step": 30530 }, { "grad_norm": 0.49419105052948, "learning_rate": 7.28117881993654e-05, "loss": 0.0304, "step": 30540 }, { "grad_norm": 0.2254517376422882, "learning_rate": 7.279339431475598e-05, "loss": 0.0237, "step": 30550 }, { "grad_norm": 0.23374193906784058, "learning_rate": 7.277499653537887e-05, "loss": 0.0206, "step": 30560 }, { "grad_norm": 0.267730176448822, "learning_rate": 7.275659486437776e-05, "loss": 0.0274, "step": 30570 }, { "grad_norm": 0.25850677490234375, "learning_rate": 7.273818930489695e-05, "loss": 0.0328, "step": 30580 }, { "grad_norm": 0.3788453936576843, "learning_rate": 7.271977986008151e-05, "loss": 0.0258, "step": 30590 }, { "grad_norm": 0.3650272786617279, "learning_rate": 7.270136653307705e-05, "loss": 0.0227, "step": 30600 }, { "grad_norm": 0.3197607398033142, "learning_rate": 7.268294932702994e-05, "loss": 0.0257, "step": 30610 }, { "grad_norm": 0.3190416991710663, "learning_rate": 7.266452824508719e-05, "loss": 0.0307, "step": 30620 }, { "grad_norm": 0.5280962586402893, "learning_rate": 7.264610329039643e-05, "loss": 0.0297, "step": 30630 }, { "grad_norm": 0.3511117100715637, "learning_rate": 7.262767446610599e-05, "loss": 0.0245, "step": 30640 }, { "grad_norm": 0.37631088495254517, "learning_rate": 7.260924177536485e-05, "loss": 0.0314, "step": 30650 }, { "grad_norm": 0.30868005752563477, "learning_rate": 7.259080522132265e-05, "loss": 0.0277, "step": 30660 }, { "grad_norm": 0.3912186622619629, "learning_rate": 7.257236480712972e-05, "loss": 0.0316, "step": 30670 }, { "grad_norm": 0.33399778604507446, "learning_rate": 7.255392053593697e-05, "loss": 0.0226, "step": 30680 }, { "grad_norm": 0.34904512763023376, "learning_rate": 7.253547241089607e-05, "loss": 0.025, "step": 30690 }, { "grad_norm": 0.6626896858215332, "learning_rate": 7.251702043515927e-05, "loss": 0.0262, "step": 30700 }, { "grad_norm": 0.2595236003398895, "learning_rate": 7.249856461187952e-05, "loss": 0.0257, "step": 30710 }, { "grad_norm": 0.3413751423358917, "learning_rate": 7.248010494421042e-05, "loss": 0.0235, "step": 30720 }, { "grad_norm": 0.3312629461288452, "learning_rate": 7.246164143530622e-05, "loss": 0.0346, "step": 30730 }, { "grad_norm": 0.24528418481349945, "learning_rate": 7.244317408832181e-05, "loss": 0.0292, "step": 30740 }, { "grad_norm": 0.3041732907295227, "learning_rate": 7.242470290641279e-05, "loss": 0.0254, "step": 30750 }, { "grad_norm": 0.36646440625190735, "learning_rate": 7.240622789273536e-05, "loss": 0.0255, "step": 30760 }, { "grad_norm": 0.35054391622543335, "learning_rate": 7.238774905044638e-05, "loss": 0.0271, "step": 30770 }, { "grad_norm": 0.30844593048095703, "learning_rate": 7.236926638270341e-05, "loss": 0.0267, "step": 30780 }, { "grad_norm": 0.3997153043746948, "learning_rate": 7.23507798926646e-05, "loss": 0.0331, "step": 30790 }, { "grad_norm": 0.284328818321228, "learning_rate": 7.23322895834888e-05, "loss": 0.0299, "step": 30800 }, { "grad_norm": 0.36867302656173706, "learning_rate": 7.231379545833552e-05, "loss": 0.0316, "step": 30810 }, { "grad_norm": 0.3199438154697418, "learning_rate": 7.229529752036487e-05, "loss": 0.0294, "step": 30820 }, { "grad_norm": 0.28098243474960327, "learning_rate": 7.227679577273765e-05, "loss": 0.0219, "step": 30830 }, { "grad_norm": 0.4557351768016815, "learning_rate": 7.225829021861529e-05, "loss": 0.0336, "step": 30840 }, { "grad_norm": 0.3284239172935486, "learning_rate": 7.223978086115992e-05, "loss": 0.0241, "step": 30850 }, { "grad_norm": 0.2758863568305969, "learning_rate": 7.222126770353425e-05, "loss": 0.0317, "step": 30860 }, { "grad_norm": 0.26772651076316833, "learning_rate": 7.22027507489017e-05, "loss": 0.0272, "step": 30870 }, { "grad_norm": 0.3775811493396759, "learning_rate": 7.218423000042627e-05, "loss": 0.0299, "step": 30880 }, { "grad_norm": 0.35461878776550293, "learning_rate": 7.216570546127268e-05, "loss": 0.0273, "step": 30890 }, { "grad_norm": 0.3078065514564514, "learning_rate": 7.214717713460626e-05, "loss": 0.0282, "step": 30900 }, { "grad_norm": 0.3805634081363678, "learning_rate": 7.2128645023593e-05, "loss": 0.0274, "step": 30910 }, { "grad_norm": 0.2420371174812317, "learning_rate": 7.211010913139951e-05, "loss": 0.0311, "step": 30920 }, { "grad_norm": 0.36308553814888, "learning_rate": 7.209156946119308e-05, "loss": 0.025, "step": 30930 }, { "grad_norm": 0.30441945791244507, "learning_rate": 7.207302601614166e-05, "loss": 0.0232, "step": 30940 }, { "grad_norm": 0.28224679827690125, "learning_rate": 7.205447879941378e-05, "loss": 0.0293, "step": 30950 }, { "grad_norm": 0.41698452830314636, "learning_rate": 7.203592781417866e-05, "loss": 0.0284, "step": 30960 }, { "grad_norm": 0.4576161503791809, "learning_rate": 7.201737306360617e-05, "loss": 0.0289, "step": 30970 }, { "grad_norm": 0.5056837201118469, "learning_rate": 7.19988145508668e-05, "loss": 0.0246, "step": 30980 }, { "grad_norm": 0.3797203302383423, "learning_rate": 7.198025227913168e-05, "loss": 0.0285, "step": 30990 }, { "grad_norm": 0.35221999883651733, "learning_rate": 7.196168625157261e-05, "loss": 0.0309, "step": 31000 }, { "grad_norm": 0.27336573600769043, "learning_rate": 7.194311647136201e-05, "loss": 0.0263, "step": 31010 }, { "grad_norm": 0.4299470782279968, "learning_rate": 7.192454294167297e-05, "loss": 0.0249, "step": 31020 }, { "grad_norm": 0.39591875672340393, "learning_rate": 7.190596566567917e-05, "loss": 0.0331, "step": 31030 }, { "grad_norm": 0.4081628918647766, "learning_rate": 7.188738464655496e-05, "loss": 0.0367, "step": 31040 }, { "grad_norm": 0.567918598651886, "learning_rate": 7.186879988747533e-05, "loss": 0.0309, "step": 31050 }, { "grad_norm": 0.3339611887931824, "learning_rate": 7.185021139161592e-05, "loss": 0.03, "step": 31060 }, { "grad_norm": 0.2678319215774536, "learning_rate": 7.1831619162153e-05, "loss": 0.0293, "step": 31070 }, { "grad_norm": 0.4390084147453308, "learning_rate": 7.181302320226345e-05, "loss": 0.0283, "step": 31080 }, { "grad_norm": 0.3260250985622406, "learning_rate": 7.179442351512482e-05, "loss": 0.031, "step": 31090 }, { "grad_norm": 0.39518919587135315, "learning_rate": 7.177582010391528e-05, "loss": 0.0314, "step": 31100 }, { "grad_norm": 0.3867904245853424, "learning_rate": 7.175721297181366e-05, "loss": 0.0378, "step": 31110 }, { "grad_norm": 0.39252036809921265, "learning_rate": 7.173860212199942e-05, "loss": 0.0314, "step": 31120 }, { "grad_norm": 0.4438939392566681, "learning_rate": 7.171998755765263e-05, "loss": 0.0295, "step": 31130 }, { "grad_norm": 0.3842918574810028, "learning_rate": 7.170136928195398e-05, "loss": 0.0252, "step": 31140 }, { "grad_norm": 0.24774384498596191, "learning_rate": 7.168274729808489e-05, "loss": 0.0219, "step": 31150 }, { "grad_norm": 0.36639147996902466, "learning_rate": 7.166412160922728e-05, "loss": 0.0307, "step": 31160 }, { "grad_norm": 0.4261327087879181, "learning_rate": 7.164549221856382e-05, "loss": 0.0283, "step": 31170 }, { "grad_norm": 0.26334476470947266, "learning_rate": 7.162685912927775e-05, "loss": 0.0417, "step": 31180 }, { "grad_norm": 0.24037271738052368, "learning_rate": 7.160822234455294e-05, "loss": 0.0237, "step": 31190 }, { "grad_norm": 0.3081531226634979, "learning_rate": 7.158958186757391e-05, "loss": 0.0272, "step": 31200 }, { "grad_norm": 0.5329634547233582, "learning_rate": 7.157093770152582e-05, "loss": 0.0326, "step": 31210 }, { "grad_norm": 0.4029550850391388, "learning_rate": 7.155228984959446e-05, "loss": 0.0294, "step": 31220 }, { "grad_norm": 0.3897201418876648, "learning_rate": 7.153363831496621e-05, "loss": 0.0346, "step": 31230 }, { "grad_norm": 0.3308725655078888, "learning_rate": 7.151498310082811e-05, "loss": 0.0398, "step": 31240 }, { "grad_norm": 0.45039284229278564, "learning_rate": 7.149632421036784e-05, "loss": 0.0346, "step": 31250 }, { "grad_norm": 0.4303690791130066, "learning_rate": 7.147766164677369e-05, "loss": 0.0467, "step": 31260 }, { "grad_norm": 0.48632293939590454, "learning_rate": 7.145899541323459e-05, "loss": 0.0315, "step": 31270 }, { "grad_norm": 0.4047054350376129, "learning_rate": 7.144032551294007e-05, "loss": 0.0378, "step": 31280 }, { "grad_norm": 0.46735045313835144, "learning_rate": 7.14216519490803e-05, "loss": 0.0374, "step": 31290 }, { "grad_norm": 0.46239280700683594, "learning_rate": 7.140297472484609e-05, "loss": 0.0414, "step": 31300 }, { "grad_norm": 0.28692981600761414, "learning_rate": 7.138429384342891e-05, "loss": 0.0425, "step": 31310 }, { "grad_norm": 0.21212337911128998, "learning_rate": 7.136560930802074e-05, "loss": 0.0275, "step": 31320 }, { "grad_norm": 0.2789137661457062, "learning_rate": 7.134692112181431e-05, "loss": 0.0272, "step": 31330 }, { "grad_norm": 0.655890941619873, "learning_rate": 7.13282292880029e-05, "loss": 0.0308, "step": 31340 }, { "grad_norm": 0.338618665933609, "learning_rate": 7.130953380978043e-05, "loss": 0.0294, "step": 31350 }, { "grad_norm": 0.762727677822113, "learning_rate": 7.129083469034144e-05, "loss": 0.0382, "step": 31360 }, { "grad_norm": 0.3946553170681, "learning_rate": 7.127213193288112e-05, "loss": 0.0278, "step": 31370 }, { "grad_norm": 0.34222880005836487, "learning_rate": 7.125342554059522e-05, "loss": 0.0262, "step": 31380 }, { "grad_norm": 0.2519218325614929, "learning_rate": 7.12347155166802e-05, "loss": 0.0307, "step": 31390 }, { "grad_norm": 0.2405262142419815, "learning_rate": 7.121600186433306e-05, "loss": 0.0238, "step": 31400 }, { "grad_norm": 0.28970667719841003, "learning_rate": 7.119728458675148e-05, "loss": 0.0372, "step": 31410 }, { "grad_norm": 0.3715522587299347, "learning_rate": 7.117856368713369e-05, "loss": 0.0295, "step": 31420 }, { "grad_norm": 0.4609137177467346, "learning_rate": 7.115983916867861e-05, "loss": 0.0327, "step": 31430 }, { "grad_norm": 0.2053307294845581, "learning_rate": 7.114111103458574e-05, "loss": 0.0236, "step": 31440 }, { "grad_norm": 0.27425941824913025, "learning_rate": 7.11223792880552e-05, "loss": 0.0256, "step": 31450 }, { "grad_norm": 0.3393137753009796, "learning_rate": 7.110364393228773e-05, "loss": 0.0293, "step": 31460 }, { "grad_norm": 0.2976994812488556, "learning_rate": 7.108490497048471e-05, "loss": 0.0258, "step": 31470 }, { "grad_norm": 0.4042441248893738, "learning_rate": 7.10661624058481e-05, "loss": 0.0311, "step": 31480 }, { "grad_norm": 0.2180624008178711, "learning_rate": 7.10474162415805e-05, "loss": 0.0279, "step": 31490 }, { "grad_norm": 0.3151831328868866, "learning_rate": 7.102866648088511e-05, "loss": 0.0301, "step": 31500 }, { "grad_norm": 0.4014187157154083, "learning_rate": 7.100991312696576e-05, "loss": 0.0296, "step": 31510 }, { "grad_norm": 0.3137114942073822, "learning_rate": 7.099115618302686e-05, "loss": 0.0353, "step": 31520 }, { "grad_norm": 0.34199491143226624, "learning_rate": 7.097239565227349e-05, "loss": 0.038, "step": 31530 }, { "grad_norm": 0.43362995982170105, "learning_rate": 7.09536315379113e-05, "loss": 0.0308, "step": 31540 }, { "grad_norm": 0.3945622444152832, "learning_rate": 7.093486384314656e-05, "loss": 0.0261, "step": 31550 }, { "grad_norm": 0.4842112064361572, "learning_rate": 7.091609257118616e-05, "loss": 0.0302, "step": 31560 }, { "grad_norm": 0.3932725191116333, "learning_rate": 7.08973177252376e-05, "loss": 0.028, "step": 31570 }, { "grad_norm": 0.32747122645378113, "learning_rate": 7.087853930850898e-05, "loss": 0.0273, "step": 31580 }, { "grad_norm": 0.2802598178386688, "learning_rate": 7.085975732420903e-05, "loss": 0.0271, "step": 31590 }, { "grad_norm": 0.2317299097776413, "learning_rate": 7.084097177554706e-05, "loss": 0.0214, "step": 31600 }, { "grad_norm": 0.2996234893798828, "learning_rate": 7.082218266573301e-05, "loss": 0.0353, "step": 31610 }, { "grad_norm": 0.39050644636154175, "learning_rate": 7.080338999797743e-05, "loss": 0.0294, "step": 31620 }, { "grad_norm": 0.28055956959724426, "learning_rate": 7.07845937754915e-05, "loss": 0.032, "step": 31630 }, { "grad_norm": 0.37561649084091187, "learning_rate": 7.076579400148693e-05, "loss": 0.0389, "step": 31640 }, { "grad_norm": 0.42859789729118347, "learning_rate": 7.074699067917611e-05, "loss": 0.032, "step": 31650 }, { "grad_norm": 0.33324941992759705, "learning_rate": 7.072818381177201e-05, "loss": 0.0282, "step": 31660 }, { "grad_norm": 0.3179995119571686, "learning_rate": 7.070937340248823e-05, "loss": 0.0254, "step": 31670 }, { "grad_norm": 0.3554513454437256, "learning_rate": 7.069055945453893e-05, "loss": 0.0226, "step": 31680 }, { "grad_norm": 0.430148184299469, "learning_rate": 7.067174197113892e-05, "loss": 0.0241, "step": 31690 }, { "grad_norm": 0.37828028202056885, "learning_rate": 7.065292095550355e-05, "loss": 0.0308, "step": 31700 }, { "grad_norm": 0.36251285672187805, "learning_rate": 7.063409641084887e-05, "loss": 0.0308, "step": 31710 }, { "grad_norm": 0.3341139256954193, "learning_rate": 7.061526834039145e-05, "loss": 0.0412, "step": 31720 }, { "grad_norm": 0.50164395570755, "learning_rate": 7.05964367473485e-05, "loss": 0.0272, "step": 31730 }, { "grad_norm": 0.26780763268470764, "learning_rate": 7.057760163493783e-05, "loss": 0.0324, "step": 31740 }, { "grad_norm": 0.39689233899116516, "learning_rate": 7.055876300637783e-05, "loss": 0.0308, "step": 31750 }, { "grad_norm": 0.31853654980659485, "learning_rate": 7.053992086488753e-05, "loss": 0.0301, "step": 31760 }, { "grad_norm": 0.3396022319793701, "learning_rate": 7.052107521368651e-05, "loss": 0.0378, "step": 31770 }, { "grad_norm": 0.3395942449569702, "learning_rate": 7.0502226055995e-05, "loss": 0.0353, "step": 31780 }, { "grad_norm": 0.33906105160713196, "learning_rate": 7.048337339503379e-05, "loss": 0.0393, "step": 31790 }, { "grad_norm": 0.42424774169921875, "learning_rate": 7.046451723402427e-05, "loss": 0.0284, "step": 31800 }, { "grad_norm": 0.34107130765914917, "learning_rate": 7.044565757618848e-05, "loss": 0.0433, "step": 31810 }, { "grad_norm": 0.33428099751472473, "learning_rate": 7.042679442474899e-05, "loss": 0.0267, "step": 31820 }, { "grad_norm": 0.2665063440799713, "learning_rate": 7.040792778292902e-05, "loss": 0.0339, "step": 31830 }, { "grad_norm": 0.332207590341568, "learning_rate": 7.038905765395234e-05, "loss": 0.0324, "step": 31840 }, { "grad_norm": 0.4355589747428894, "learning_rate": 7.037018404104334e-05, "loss": 0.0352, "step": 31850 }, { "grad_norm": 0.3364366888999939, "learning_rate": 7.035130694742702e-05, "loss": 0.0275, "step": 31860 }, { "grad_norm": 0.3463004529476166, "learning_rate": 7.033242637632897e-05, "loss": 0.0307, "step": 31870 }, { "grad_norm": 0.3144301772117615, "learning_rate": 7.031354233097534e-05, "loss": 0.0285, "step": 31880 }, { "grad_norm": 0.2637016773223877, "learning_rate": 7.029465481459289e-05, "loss": 0.0374, "step": 31890 }, { "grad_norm": 0.2259361743927002, "learning_rate": 7.027576383040898e-05, "loss": 0.024, "step": 31900 }, { "grad_norm": 0.32909393310546875, "learning_rate": 7.025686938165159e-05, "loss": 0.0299, "step": 31910 }, { "grad_norm": 0.3472478687763214, "learning_rate": 7.023797147154924e-05, "loss": 0.035, "step": 31920 }, { "grad_norm": 0.38205647468566895, "learning_rate": 7.021907010333111e-05, "loss": 0.0247, "step": 31930 }, { "grad_norm": 0.2781842052936554, "learning_rate": 7.020016528022685e-05, "loss": 0.0223, "step": 31940 }, { "grad_norm": 0.4218674302101135, "learning_rate": 7.018125700546683e-05, "loss": 0.0308, "step": 31950 }, { "grad_norm": 0.29958558082580566, "learning_rate": 7.016234528228196e-05, "loss": 0.0337, "step": 31960 }, { "grad_norm": 0.42303892970085144, "learning_rate": 7.014343011390372e-05, "loss": 0.0286, "step": 31970 }, { "grad_norm": 0.2149171084165573, "learning_rate": 7.01245115035642e-05, "loss": 0.0195, "step": 31980 }, { "grad_norm": 0.36685436964035034, "learning_rate": 7.010558945449606e-05, "loss": 0.0372, "step": 31990 }, { "grad_norm": 0.2984188497066498, "learning_rate": 7.008666396993258e-05, "loss": 0.0232, "step": 32000 }, { "grad_norm": 0.4626922905445099, "learning_rate": 7.006773505310759e-05, "loss": 0.0326, "step": 32010 }, { "grad_norm": 0.2850149869918823, "learning_rate": 7.004880270725553e-05, "loss": 0.0297, "step": 32020 }, { "grad_norm": 0.30719178915023804, "learning_rate": 7.002986693561144e-05, "loss": 0.027, "step": 32030 }, { "grad_norm": 0.4062868654727936, "learning_rate": 7.001092774141089e-05, "loss": 0.0309, "step": 32040 }, { "grad_norm": 0.34291061758995056, "learning_rate": 6.999198512789009e-05, "loss": 0.0278, "step": 32050 }, { "grad_norm": 0.34300583600997925, "learning_rate": 6.997303909828584e-05, "loss": 0.0273, "step": 32060 }, { "grad_norm": 0.4434550106525421, "learning_rate": 6.995408965583544e-05, "loss": 0.0269, "step": 32070 }, { "grad_norm": 0.2568579316139221, "learning_rate": 6.993513680377688e-05, "loss": 0.0198, "step": 32080 }, { "grad_norm": 0.2672926187515259, "learning_rate": 6.991618054534868e-05, "loss": 0.0354, "step": 32090 }, { "grad_norm": 0.496101438999176, "learning_rate": 6.989722088378991e-05, "loss": 0.021, "step": 32100 }, { "grad_norm": 0.5306718349456787, "learning_rate": 6.987825782234027e-05, "loss": 0.0319, "step": 32110 }, { "grad_norm": 0.27743151783943176, "learning_rate": 6.985929136424006e-05, "loss": 0.023, "step": 32120 }, { "grad_norm": 0.34809231758117676, "learning_rate": 6.984032151273012e-05, "loss": 0.0255, "step": 32130 }, { "grad_norm": 0.4674510657787323, "learning_rate": 6.982134827105186e-05, "loss": 0.0295, "step": 32140 }, { "grad_norm": 0.35652270913124084, "learning_rate": 6.980237164244729e-05, "loss": 0.0266, "step": 32150 }, { "grad_norm": 0.3297632038593292, "learning_rate": 6.9783391630159e-05, "loss": 0.0262, "step": 32160 }, { "grad_norm": 0.4282466471195221, "learning_rate": 6.976440823743015e-05, "loss": 0.0281, "step": 32170 }, { "grad_norm": 0.28919416666030884, "learning_rate": 6.974542146750451e-05, "loss": 0.0231, "step": 32180 }, { "grad_norm": 0.3301655352115631, "learning_rate": 6.972643132362637e-05, "loss": 0.0226, "step": 32190 }, { "grad_norm": 0.3169601559638977, "learning_rate": 6.970743780904064e-05, "loss": 0.0252, "step": 32200 }, { "grad_norm": 0.3576303720474243, "learning_rate": 6.968844092699277e-05, "loss": 0.0261, "step": 32210 }, { "grad_norm": 0.31123071908950806, "learning_rate": 6.966944068072883e-05, "loss": 0.0227, "step": 32220 }, { "grad_norm": 0.3329331874847412, "learning_rate": 6.965043707349545e-05, "loss": 0.0229, "step": 32230 }, { "grad_norm": 0.6596238613128662, "learning_rate": 6.963143010853982e-05, "loss": 0.0332, "step": 32240 }, { "grad_norm": 0.5643609762191772, "learning_rate": 6.961241978910971e-05, "loss": 0.0319, "step": 32250 }, { "grad_norm": 0.5023745894432068, "learning_rate": 6.959340611845344e-05, "loss": 0.0252, "step": 32260 }, { "grad_norm": 0.3982876241207123, "learning_rate": 6.957438909981995e-05, "loss": 0.0276, "step": 32270 }, { "grad_norm": 0.3091343939304352, "learning_rate": 6.955536873645872e-05, "loss": 0.0244, "step": 32280 }, { "grad_norm": 0.3948211669921875, "learning_rate": 6.953634503161982e-05, "loss": 0.0315, "step": 32290 }, { "grad_norm": 0.2966434955596924, "learning_rate": 6.951731798855387e-05, "loss": 0.0261, "step": 32300 }, { "grad_norm": 0.29150670766830444, "learning_rate": 6.949828761051208e-05, "loss": 0.0347, "step": 32310 }, { "grad_norm": 0.34524041414260864, "learning_rate": 6.947925390074622e-05, "loss": 0.0213, "step": 32320 }, { "grad_norm": 0.2969055473804474, "learning_rate": 6.946021686250863e-05, "loss": 0.0258, "step": 32330 }, { "grad_norm": 0.3510313630104065, "learning_rate": 6.94411764990522e-05, "loss": 0.0254, "step": 32340 }, { "grad_norm": 0.3106897473335266, "learning_rate": 6.942213281363044e-05, "loss": 0.0233, "step": 32350 }, { "grad_norm": 0.35239166021347046, "learning_rate": 6.940308580949737e-05, "loss": 0.0284, "step": 32360 }, { "grad_norm": 0.19377762079238892, "learning_rate": 6.93840354899076e-05, "loss": 0.0247, "step": 32370 }, { "grad_norm": 0.381287544965744, "learning_rate": 6.936498185811633e-05, "loss": 0.0308, "step": 32380 }, { "grad_norm": 0.20671352744102478, "learning_rate": 6.93459249173793e-05, "loss": 0.0212, "step": 32390 }, { "grad_norm": 0.5012876391410828, "learning_rate": 6.932686467095279e-05, "loss": 0.0303, "step": 32400 }, { "grad_norm": 0.3751028776168823, "learning_rate": 6.930780112209373e-05, "loss": 0.0265, "step": 32410 }, { "grad_norm": 0.4597599506378174, "learning_rate": 6.92887342740595e-05, "loss": 0.0315, "step": 32420 }, { "grad_norm": 0.27642393112182617, "learning_rate": 6.926966413010816e-05, "loss": 0.0307, "step": 32430 }, { "grad_norm": 0.5281280279159546, "learning_rate": 6.925059069349824e-05, "loss": 0.0232, "step": 32440 }, { "grad_norm": 0.7906391620635986, "learning_rate": 6.923151396748886e-05, "loss": 0.0301, "step": 32450 }, { "grad_norm": 0.3577786684036255, "learning_rate": 6.921243395533974e-05, "loss": 0.0347, "step": 32460 }, { "grad_norm": 0.42104262113571167, "learning_rate": 6.919335066031109e-05, "loss": 0.0367, "step": 32470 }, { "grad_norm": 0.46347326040267944, "learning_rate": 6.917426408566379e-05, "loss": 0.0393, "step": 32480 }, { "grad_norm": 0.42334121465682983, "learning_rate": 6.915517423465916e-05, "loss": 0.0278, "step": 32490 }, { "grad_norm": 0.3060961365699768, "learning_rate": 6.913608111055914e-05, "loss": 0.0269, "step": 32500 }, { "grad_norm": 0.22703160345554352, "learning_rate": 6.911698471662623e-05, "loss": 0.0225, "step": 32510 }, { "grad_norm": 0.23628169298171997, "learning_rate": 6.90978850561235e-05, "loss": 0.024, "step": 32520 }, { "grad_norm": 0.28979647159576416, "learning_rate": 6.907878213231454e-05, "loss": 0.0273, "step": 32530 }, { "grad_norm": 0.4375649690628052, "learning_rate": 6.90596759484635e-05, "loss": 0.0268, "step": 32540 }, { "grad_norm": 0.3913542628288269, "learning_rate": 6.904056650783514e-05, "loss": 0.029, "step": 32550 }, { "grad_norm": 0.44816797971725464, "learning_rate": 6.902145381369471e-05, "loss": 0.0308, "step": 32560 }, { "grad_norm": 1.6080936193466187, "learning_rate": 6.900233786930808e-05, "loss": 0.0342, "step": 32570 }, { "grad_norm": 0.3665081560611725, "learning_rate": 6.898321867794161e-05, "loss": 0.0326, "step": 32580 }, { "grad_norm": 0.23146283626556396, "learning_rate": 6.896409624286226e-05, "loss": 0.0261, "step": 32590 }, { "grad_norm": 0.34659263491630554, "learning_rate": 6.894497056733754e-05, "loss": 0.0352, "step": 32600 }, { "grad_norm": 0.315106064081192, "learning_rate": 6.89258416546355e-05, "loss": 0.0268, "step": 32610 }, { "grad_norm": 0.34644460678100586, "learning_rate": 6.890670950802474e-05, "loss": 0.0286, "step": 32620 }, { "grad_norm": 0.2622115910053253, "learning_rate": 6.88875741307744e-05, "loss": 0.0306, "step": 32630 }, { "grad_norm": 0.25413843989372253, "learning_rate": 6.886843552615425e-05, "loss": 0.0269, "step": 32640 }, { "grad_norm": 0.4047396183013916, "learning_rate": 6.884929369743451e-05, "loss": 0.0316, "step": 32650 }, { "grad_norm": 0.47431451082229614, "learning_rate": 6.8830148647886e-05, "loss": 0.0332, "step": 32660 }, { "grad_norm": 0.2795354723930359, "learning_rate": 6.88110003807801e-05, "loss": 0.038, "step": 32670 }, { "grad_norm": 0.3602047860622406, "learning_rate": 6.87918488993887e-05, "loss": 0.0288, "step": 32680 }, { "grad_norm": 0.4977492094039917, "learning_rate": 6.877269420698431e-05, "loss": 0.0319, "step": 32690 }, { "grad_norm": 0.5328190326690674, "learning_rate": 6.875353630683989e-05, "loss": 0.0306, "step": 32700 }, { "grad_norm": 0.6245599389076233, "learning_rate": 6.873437520222905e-05, "loss": 0.035, "step": 32710 }, { "grad_norm": 0.3385064899921417, "learning_rate": 6.871521089642585e-05, "loss": 0.028, "step": 32720 }, { "grad_norm": 0.480318158864975, "learning_rate": 6.869604339270498e-05, "loss": 0.0301, "step": 32730 }, { "grad_norm": 0.3335191607475281, "learning_rate": 6.867687269434164e-05, "loss": 0.0359, "step": 32740 }, { "grad_norm": 0.23747995495796204, "learning_rate": 6.865769880461156e-05, "loss": 0.0293, "step": 32750 }, { "grad_norm": 0.4301336407661438, "learning_rate": 6.863852172679104e-05, "loss": 0.0281, "step": 32760 }, { "grad_norm": 0.3401126563549042, "learning_rate": 6.861934146415693e-05, "loss": 0.0279, "step": 32770 }, { "grad_norm": 0.32493850588798523, "learning_rate": 6.86001580199866e-05, "loss": 0.0324, "step": 32780 }, { "grad_norm": 0.28920239210128784, "learning_rate": 6.858097139755798e-05, "loss": 0.0259, "step": 32790 }, { "grad_norm": 0.41564980149269104, "learning_rate": 6.856178160014955e-05, "loss": 0.0279, "step": 32800 }, { "grad_norm": 0.32094916701316833, "learning_rate": 6.85425886310403e-05, "loss": 0.029, "step": 32810 }, { "grad_norm": 0.402759850025177, "learning_rate": 6.852339249350979e-05, "loss": 0.0346, "step": 32820 }, { "grad_norm": 0.42448875308036804, "learning_rate": 6.850419319083812e-05, "loss": 0.0356, "step": 32830 }, { "grad_norm": 0.36204588413238525, "learning_rate": 6.848499072630592e-05, "loss": 0.0198, "step": 32840 }, { "grad_norm": 0.34575387835502625, "learning_rate": 6.846578510319439e-05, "loss": 0.034, "step": 32850 }, { "grad_norm": 0.2894441485404968, "learning_rate": 6.844657632478519e-05, "loss": 0.0351, "step": 32860 }, { "grad_norm": 0.2770766019821167, "learning_rate": 6.842736439436063e-05, "loss": 0.0222, "step": 32870 }, { "grad_norm": 0.28932932019233704, "learning_rate": 6.84081493152035e-05, "loss": 0.0237, "step": 32880 }, { "grad_norm": 0.33409205079078674, "learning_rate": 6.83889310905971e-05, "loss": 0.023, "step": 32890 }, { "grad_norm": 0.31593620777130127, "learning_rate": 6.836970972382533e-05, "loss": 0.0345, "step": 32900 }, { "grad_norm": 0.26580891013145447, "learning_rate": 6.835048521817257e-05, "loss": 0.0343, "step": 32910 }, { "grad_norm": 0.3802926242351532, "learning_rate": 6.833125757692379e-05, "loss": 0.0287, "step": 32920 }, { "grad_norm": 0.4446367919445038, "learning_rate": 6.831202680336441e-05, "loss": 0.0263, "step": 32930 }, { "grad_norm": 0.24659128487110138, "learning_rate": 6.829279290078052e-05, "loss": 0.0276, "step": 32940 }, { "grad_norm": 0.2750090956687927, "learning_rate": 6.827355587245863e-05, "loss": 0.0295, "step": 32950 }, { "grad_norm": 0.439386248588562, "learning_rate": 6.82543157216858e-05, "loss": 0.024, "step": 32960 }, { "grad_norm": 0.19932830333709717, "learning_rate": 6.823507245174969e-05, "loss": 0.0208, "step": 32970 }, { "grad_norm": 0.24769514799118042, "learning_rate": 6.821582606593841e-05, "loss": 0.0302, "step": 32980 }, { "grad_norm": 0.36618027091026306, "learning_rate": 6.81965765675407e-05, "loss": 0.0336, "step": 32990 }, { "grad_norm": 0.32095757126808167, "learning_rate": 6.81773239598457e-05, "loss": 0.0338, "step": 33000 }, { "grad_norm": 0.43100786209106445, "learning_rate": 6.815806824614319e-05, "loss": 0.032, "step": 33010 }, { "grad_norm": 0.27785220742225647, "learning_rate": 6.813880942972343e-05, "loss": 0.0224, "step": 33020 }, { "grad_norm": 0.42583373188972473, "learning_rate": 6.811954751387726e-05, "loss": 0.0296, "step": 33030 }, { "grad_norm": 0.2553226351737976, "learning_rate": 6.810028250189598e-05, "loss": 0.0197, "step": 33040 }, { "grad_norm": 0.3659006953239441, "learning_rate": 6.808101439707147e-05, "loss": 0.0344, "step": 33050 }, { "grad_norm": 0.3879025876522064, "learning_rate": 6.806174320269609e-05, "loss": 0.0321, "step": 33060 }, { "grad_norm": 0.34536606073379517, "learning_rate": 6.804246892206281e-05, "loss": 0.0331, "step": 33070 }, { "grad_norm": 0.3015308380126953, "learning_rate": 6.802319155846506e-05, "loss": 0.041, "step": 33080 }, { "grad_norm": 0.31123313307762146, "learning_rate": 6.800391111519679e-05, "loss": 0.0274, "step": 33090 }, { "grad_norm": 0.48910319805145264, "learning_rate": 6.798462759555253e-05, "loss": 0.0297, "step": 33100 }, { "grad_norm": 0.23475754261016846, "learning_rate": 6.79653410028273e-05, "loss": 0.0265, "step": 33110 }, { "grad_norm": 0.3337107300758362, "learning_rate": 6.794605134031663e-05, "loss": 0.03, "step": 33120 }, { "grad_norm": 0.35441702604293823, "learning_rate": 6.792675861131661e-05, "loss": 0.0292, "step": 33130 }, { "grad_norm": 0.344229519367218, "learning_rate": 6.790746281912386e-05, "loss": 0.0322, "step": 33140 }, { "grad_norm": 0.31228891015052795, "learning_rate": 6.788816396703546e-05, "loss": 0.0241, "step": 33150 }, { "grad_norm": 0.37969619035720825, "learning_rate": 6.78688620583491e-05, "loss": 0.0285, "step": 33160 }, { "grad_norm": 0.3606884181499481, "learning_rate": 6.784955709636292e-05, "loss": 0.0329, "step": 33170 }, { "grad_norm": 0.5573069453239441, "learning_rate": 6.783024908437564e-05, "loss": 0.0366, "step": 33180 }, { "grad_norm": 0.2738828659057617, "learning_rate": 6.781093802568641e-05, "loss": 0.0298, "step": 33190 }, { "grad_norm": 0.3277484178543091, "learning_rate": 6.779162392359504e-05, "loss": 0.0309, "step": 33200 }, { "grad_norm": 0.4910268187522888, "learning_rate": 6.777230678140172e-05, "loss": 0.0349, "step": 33210 }, { "grad_norm": 0.26097404956817627, "learning_rate": 6.775298660240726e-05, "loss": 0.0322, "step": 33220 }, { "grad_norm": 0.3385092318058014, "learning_rate": 6.773366338991292e-05, "loss": 0.0405, "step": 33230 }, { "grad_norm": 0.4251457452774048, "learning_rate": 6.771433714722052e-05, "loss": 0.0275, "step": 33240 }, { "grad_norm": 0.6067336797714233, "learning_rate": 6.769500787763239e-05, "loss": 0.0339, "step": 33250 }, { "grad_norm": 0.38397446274757385, "learning_rate": 6.76756755844514e-05, "loss": 0.0321, "step": 33260 }, { "grad_norm": 0.2715092599391937, "learning_rate": 6.765634027098087e-05, "loss": 0.0273, "step": 33270 }, { "grad_norm": 0.42344650626182556, "learning_rate": 6.763700194052468e-05, "loss": 0.0338, "step": 33280 }, { "grad_norm": 0.4853442907333374, "learning_rate": 6.761766059638723e-05, "loss": 0.0269, "step": 33290 }, { "grad_norm": 0.3363347053527832, "learning_rate": 6.759831624187345e-05, "loss": 0.0255, "step": 33300 }, { "grad_norm": 0.3192463517189026, "learning_rate": 6.757896888028871e-05, "loss": 0.0247, "step": 33310 }, { "grad_norm": 0.3063444197177887, "learning_rate": 6.7559618514939e-05, "loss": 0.0264, "step": 33320 }, { "grad_norm": 0.31313955783843994, "learning_rate": 6.754026514913073e-05, "loss": 0.0297, "step": 33330 }, { "grad_norm": 0.36732956767082214, "learning_rate": 6.752090878617087e-05, "loss": 0.0413, "step": 33340 }, { "grad_norm": 0.3814288377761841, "learning_rate": 6.75015494293669e-05, "loss": 0.0294, "step": 33350 }, { "grad_norm": 0.3148014545440674, "learning_rate": 6.74821870820268e-05, "loss": 0.0241, "step": 33360 }, { "grad_norm": 0.2706618905067444, "learning_rate": 6.746282174745907e-05, "loss": 0.0371, "step": 33370 }, { "grad_norm": 0.34125834703445435, "learning_rate": 6.744345342897271e-05, "loss": 0.0351, "step": 33380 }, { "grad_norm": 0.4890277087688446, "learning_rate": 6.742408212987724e-05, "loss": 0.0384, "step": 33390 }, { "grad_norm": 0.401520699262619, "learning_rate": 6.740470785348269e-05, "loss": 0.024, "step": 33400 }, { "grad_norm": 0.2824762463569641, "learning_rate": 6.738533060309958e-05, "loss": 0.0255, "step": 33410 }, { "grad_norm": 0.38704192638397217, "learning_rate": 6.736595038203894e-05, "loss": 0.0263, "step": 33420 }, { "grad_norm": 0.27216389775276184, "learning_rate": 6.734656719361236e-05, "loss": 0.0281, "step": 33430 }, { "grad_norm": 0.30627620220184326, "learning_rate": 6.732718104113189e-05, "loss": 0.0288, "step": 33440 }, { "grad_norm": 0.3121415376663208, "learning_rate": 6.730779192791006e-05, "loss": 0.0195, "step": 33450 }, { "grad_norm": 0.31791403889656067, "learning_rate": 6.728839985725997e-05, "loss": 0.0252, "step": 33460 }, { "grad_norm": 0.30851009488105774, "learning_rate": 6.726900483249517e-05, "loss": 0.0286, "step": 33470 }, { "grad_norm": 0.5301551818847656, "learning_rate": 6.724960685692976e-05, "loss": 0.0409, "step": 33480 }, { "grad_norm": 0.4397796094417572, "learning_rate": 6.723020593387833e-05, "loss": 0.029, "step": 33490 }, { "grad_norm": 0.4012430012226105, "learning_rate": 6.721080206665593e-05, "loss": 0.0297, "step": 33500 }, { "grad_norm": 0.22100959718227386, "learning_rate": 6.719139525857819e-05, "loss": 0.03, "step": 33510 }, { "grad_norm": 0.36977019906044006, "learning_rate": 6.717198551296117e-05, "loss": 0.0293, "step": 33520 }, { "grad_norm": 0.34679460525512695, "learning_rate": 6.715257283312148e-05, "loss": 0.0321, "step": 33530 }, { "grad_norm": 0.4284573793411255, "learning_rate": 6.713315722237623e-05, "loss": 0.0305, "step": 33540 }, { "grad_norm": 0.35654813051223755, "learning_rate": 6.7113738684043e-05, "loss": 0.0329, "step": 33550 }, { "grad_norm": 0.3832228481769562, "learning_rate": 6.709431722143989e-05, "loss": 0.0283, "step": 33560 }, { "grad_norm": 0.42988696694374084, "learning_rate": 6.70748928378855e-05, "loss": 0.028, "step": 33570 }, { "grad_norm": 0.23525074124336243, "learning_rate": 6.705546553669891e-05, "loss": 0.0203, "step": 33580 }, { "grad_norm": 0.21356919407844543, "learning_rate": 6.703603532119974e-05, "loss": 0.0213, "step": 33590 }, { "grad_norm": 0.26142820715904236, "learning_rate": 6.701660219470808e-05, "loss": 0.0208, "step": 33600 }, { "grad_norm": 0.4317372143268585, "learning_rate": 6.69971661605445e-05, "loss": 0.0233, "step": 33610 }, { "grad_norm": 0.3807433843612671, "learning_rate": 6.697772722203008e-05, "loss": 0.0287, "step": 33620 }, { "grad_norm": 0.30584579706192017, "learning_rate": 6.695828538248643e-05, "loss": 0.0268, "step": 33630 }, { "grad_norm": 0.5007081031799316, "learning_rate": 6.693884064523563e-05, "loss": 0.0266, "step": 33640 }, { "grad_norm": 0.2867153286933899, "learning_rate": 6.691939301360023e-05, "loss": 0.0299, "step": 33650 }, { "grad_norm": 0.2707582116127014, "learning_rate": 6.689994249090333e-05, "loss": 0.0255, "step": 33660 }, { "grad_norm": 0.48697221279144287, "learning_rate": 6.688048908046845e-05, "loss": 0.035, "step": 33670 }, { "grad_norm": 0.29652196168899536, "learning_rate": 6.686103278561969e-05, "loss": 0.0243, "step": 33680 }, { "grad_norm": 0.35209307074546814, "learning_rate": 6.684157360968156e-05, "loss": 0.0254, "step": 33690 }, { "grad_norm": 0.28143519163131714, "learning_rate": 6.682211155597911e-05, "loss": 0.0219, "step": 33700 }, { "grad_norm": 0.30014657974243164, "learning_rate": 6.680264662783789e-05, "loss": 0.0263, "step": 33710 }, { "grad_norm": 0.3727972209453583, "learning_rate": 6.678317882858391e-05, "loss": 0.0294, "step": 33720 }, { "grad_norm": 0.3109005391597748, "learning_rate": 6.67637081615437e-05, "loss": 0.0214, "step": 33730 }, { "grad_norm": 0.23554733395576477, "learning_rate": 6.674423463004427e-05, "loss": 0.0189, "step": 33740 }, { "grad_norm": 0.33440956473350525, "learning_rate": 6.672475823741308e-05, "loss": 0.0232, "step": 33750 }, { "grad_norm": 0.3632369041442871, "learning_rate": 6.670527898697811e-05, "loss": 0.0272, "step": 33760 }, { "grad_norm": 0.31421083211898804, "learning_rate": 6.668579688206788e-05, "loss": 0.0274, "step": 33770 }, { "grad_norm": 0.48725706338882446, "learning_rate": 6.666631192601131e-05, "loss": 0.0258, "step": 33780 }, { "grad_norm": 0.5076538324356079, "learning_rate": 6.664682412213785e-05, "loss": 0.0296, "step": 33790 }, { "grad_norm": 0.38163167238235474, "learning_rate": 6.662733347377745e-05, "loss": 0.0285, "step": 33800 }, { "grad_norm": 0.3052188754081726, "learning_rate": 6.660783998426051e-05, "loss": 0.0234, "step": 33810 }, { "grad_norm": 0.34123149514198303, "learning_rate": 6.658834365691794e-05, "loss": 0.0249, "step": 33820 }, { "grad_norm": 0.2592824101448059, "learning_rate": 6.656884449508115e-05, "loss": 0.0208, "step": 33830 }, { "grad_norm": 0.2142525017261505, "learning_rate": 6.654934250208198e-05, "loss": 0.0268, "step": 33840 }, { "grad_norm": 0.2564958333969116, "learning_rate": 6.65298376812528e-05, "loss": 0.0289, "step": 33850 }, { "grad_norm": 0.32597360014915466, "learning_rate": 6.651033003592646e-05, "loss": 0.031, "step": 33860 }, { "grad_norm": 0.254606693983078, "learning_rate": 6.649081956943626e-05, "loss": 0.0259, "step": 33870 }, { "grad_norm": 0.45020127296447754, "learning_rate": 6.647130628511604e-05, "loss": 0.0325, "step": 33880 }, { "grad_norm": 0.3137882351875305, "learning_rate": 6.645179018630005e-05, "loss": 0.0326, "step": 33890 }, { "grad_norm": 0.21836349368095398, "learning_rate": 6.643227127632309e-05, "loss": 0.0275, "step": 33900 }, { "grad_norm": 0.35403719544410706, "learning_rate": 6.641274955852038e-05, "loss": 0.0244, "step": 33910 }, { "grad_norm": 0.42522621154785156, "learning_rate": 6.639322503622768e-05, "loss": 0.0406, "step": 33920 }, { "grad_norm": 0.3949424922466278, "learning_rate": 6.637369771278116e-05, "loss": 0.029, "step": 33930 }, { "grad_norm": 0.5192767381668091, "learning_rate": 6.635416759151751e-05, "loss": 0.0203, "step": 33940 }, { "grad_norm": 0.344089537858963, "learning_rate": 6.633463467577394e-05, "loss": 0.0439, "step": 33950 }, { "grad_norm": 0.3205130398273468, "learning_rate": 6.631509896888803e-05, "loss": 0.0307, "step": 33960 }, { "grad_norm": 0.20283475518226624, "learning_rate": 6.629556047419794e-05, "loss": 0.0248, "step": 33970 }, { "grad_norm": 0.29626962542533875, "learning_rate": 6.627601919504223e-05, "loss": 0.0254, "step": 33980 }, { "grad_norm": 0.3673174977302551, "learning_rate": 6.625647513476001e-05, "loss": 0.021, "step": 33990 }, { "grad_norm": 0.4226592481136322, "learning_rate": 6.62369282966908e-05, "loss": 0.0363, "step": 34000 }, { "grad_norm": 0.3082541227340698, "learning_rate": 6.621737868417464e-05, "loss": 0.0309, "step": 34010 }, { "grad_norm": 0.3426045775413513, "learning_rate": 6.619782630055198e-05, "loss": 0.0285, "step": 34020 }, { "grad_norm": 0.3740784823894501, "learning_rate": 6.617827114916382e-05, "loss": 0.0273, "step": 34030 }, { "grad_norm": 0.286284863948822, "learning_rate": 6.615871323335161e-05, "loss": 0.0275, "step": 34040 }, { "grad_norm": 0.3641044497489929, "learning_rate": 6.613915255645725e-05, "loss": 0.0301, "step": 34050 }, { "grad_norm": 0.31879764795303345, "learning_rate": 6.611958912182312e-05, "loss": 0.0318, "step": 34060 }, { "grad_norm": 0.2871643006801605, "learning_rate": 6.610002293279207e-05, "loss": 0.026, "step": 34070 }, { "grad_norm": 0.2652439773082733, "learning_rate": 6.608045399270746e-05, "loss": 0.0247, "step": 34080 }, { "grad_norm": 0.2767874300479889, "learning_rate": 6.606088230491304e-05, "loss": 0.022, "step": 34090 }, { "grad_norm": 0.36073198914527893, "learning_rate": 6.604130787275312e-05, "loss": 0.023, "step": 34100 }, { "grad_norm": 0.29136455059051514, "learning_rate": 6.602173069957242e-05, "loss": 0.0278, "step": 34110 }, { "grad_norm": 0.4442812502384186, "learning_rate": 6.600215078871612e-05, "loss": 0.0246, "step": 34120 }, { "grad_norm": 0.30800023674964905, "learning_rate": 6.598256814352992e-05, "loss": 0.0254, "step": 34130 }, { "grad_norm": 0.3261723816394806, "learning_rate": 6.596298276735995e-05, "loss": 0.0211, "step": 34140 }, { "grad_norm": 0.32426026463508606, "learning_rate": 6.594339466355282e-05, "loss": 0.0277, "step": 34150 }, { "grad_norm": 0.4465731382369995, "learning_rate": 6.592380383545558e-05, "loss": 0.0218, "step": 34160 }, { "grad_norm": 0.3318762481212616, "learning_rate": 6.590421028641577e-05, "loss": 0.0328, "step": 34170 }, { "grad_norm": 0.3095495402812958, "learning_rate": 6.588461401978143e-05, "loss": 0.0287, "step": 34180 }, { "grad_norm": 0.32981225848197937, "learning_rate": 6.586501503890099e-05, "loss": 0.0285, "step": 34190 }, { "grad_norm": 0.4020749032497406, "learning_rate": 6.584541334712338e-05, "loss": 0.0311, "step": 34200 }, { "grad_norm": 0.37785810232162476, "learning_rate": 6.582580894779802e-05, "loss": 0.0288, "step": 34210 }, { "grad_norm": 0.2675769031047821, "learning_rate": 6.580620184427473e-05, "loss": 0.0228, "step": 34220 }, { "grad_norm": 0.24966482818126678, "learning_rate": 6.578659203990385e-05, "loss": 0.0238, "step": 34230 }, { "grad_norm": 0.30722957849502563, "learning_rate": 6.576697953803615e-05, "loss": 0.0193, "step": 34240 }, { "grad_norm": 0.3078114092350006, "learning_rate": 6.57473643420229e-05, "loss": 0.021, "step": 34250 }, { "grad_norm": 0.3368714153766632, "learning_rate": 6.572774645521574e-05, "loss": 0.0264, "step": 34260 }, { "grad_norm": 0.5768164992332458, "learning_rate": 6.570812588096688e-05, "loss": 0.0295, "step": 34270 }, { "grad_norm": 0.22686071693897247, "learning_rate": 6.568850262262893e-05, "loss": 0.028, "step": 34280 }, { "grad_norm": 0.3696900010108948, "learning_rate": 6.566887668355497e-05, "loss": 0.0244, "step": 34290 }, { "grad_norm": 0.381058931350708, "learning_rate": 6.564924806709851e-05, "loss": 0.0234, "step": 34300 }, { "grad_norm": 0.3771689534187317, "learning_rate": 6.562961677661359e-05, "loss": 0.0311, "step": 34310 }, { "grad_norm": 0.4413069188594818, "learning_rate": 6.56099828154546e-05, "loss": 0.0288, "step": 34320 }, { "grad_norm": 0.44834867119789124, "learning_rate": 6.55903461869765e-05, "loss": 0.0273, "step": 34330 }, { "grad_norm": 0.2769179344177246, "learning_rate": 6.557070689453465e-05, "loss": 0.0234, "step": 34340 }, { "grad_norm": 0.2857648730278015, "learning_rate": 6.555106494148482e-05, "loss": 0.0299, "step": 34350 }, { "grad_norm": 0.367144376039505, "learning_rate": 6.553142033118333e-05, "loss": 0.0255, "step": 34360 }, { "grad_norm": 0.39481881260871887, "learning_rate": 6.551177306698688e-05, "loss": 0.029, "step": 34370 }, { "grad_norm": 0.2905380129814148, "learning_rate": 6.549212315225267e-05, "loss": 0.0334, "step": 34380 }, { "grad_norm": 0.3089408576488495, "learning_rate": 6.547247059033833e-05, "loss": 0.0323, "step": 34390 }, { "grad_norm": 0.3068016767501831, "learning_rate": 6.545281538460193e-05, "loss": 0.0294, "step": 34400 }, { "grad_norm": 0.31358450651168823, "learning_rate": 6.543315753840202e-05, "loss": 0.0259, "step": 34410 }, { "grad_norm": 0.4354475140571594, "learning_rate": 6.541349705509758e-05, "loss": 0.0239, "step": 34420 }, { "grad_norm": 0.34617844223976135, "learning_rate": 6.539383393804805e-05, "loss": 0.0277, "step": 34430 }, { "grad_norm": 0.32155802845954895, "learning_rate": 6.537416819061333e-05, "loss": 0.0304, "step": 34440 }, { "grad_norm": 0.4293826222419739, "learning_rate": 6.535449981615375e-05, "loss": 0.0262, "step": 34450 }, { "grad_norm": 0.3063431680202484, "learning_rate": 6.53348288180301e-05, "loss": 0.0312, "step": 34460 }, { "grad_norm": 0.36924979090690613, "learning_rate": 6.531515519960361e-05, "loss": 0.0383, "step": 34470 }, { "grad_norm": 0.3342404365539551, "learning_rate": 6.529547896423597e-05, "loss": 0.0293, "step": 34480 }, { "grad_norm": 0.2729114294052124, "learning_rate": 6.52758001152893e-05, "loss": 0.0265, "step": 34490 }, { "grad_norm": 0.29085034132003784, "learning_rate": 6.525611865612618e-05, "loss": 0.0225, "step": 34500 }, { "grad_norm": 0.37240853905677795, "learning_rate": 6.523643459010966e-05, "loss": 0.0264, "step": 34510 }, { "grad_norm": 0.4792536497116089, "learning_rate": 6.521674792060317e-05, "loss": 0.0261, "step": 34520 }, { "grad_norm": 0.3742888271808624, "learning_rate": 6.519705865097063e-05, "loss": 0.0293, "step": 34530 }, { "grad_norm": 0.4331352710723877, "learning_rate": 6.517736678457641e-05, "loss": 0.0237, "step": 34540 }, { "grad_norm": 0.3695802092552185, "learning_rate": 6.515767232478534e-05, "loss": 0.0264, "step": 34550 }, { "grad_norm": 0.3434814214706421, "learning_rate": 6.51379752749626e-05, "loss": 0.021, "step": 34560 }, { "grad_norm": 0.28790050745010376, "learning_rate": 6.511827563847393e-05, "loss": 0.0221, "step": 34570 }, { "grad_norm": 0.3643239438533783, "learning_rate": 6.509857341868542e-05, "loss": 0.0307, "step": 34580 }, { "grad_norm": 0.34451696276664734, "learning_rate": 6.507886861896367e-05, "loss": 0.0318, "step": 34590 }, { "grad_norm": 0.45658132433891296, "learning_rate": 6.505916124267567e-05, "loss": 0.036, "step": 34600 }, { "grad_norm": 0.6887015104293823, "learning_rate": 6.503945129318891e-05, "loss": 0.0291, "step": 34610 }, { "grad_norm": 0.35739871859550476, "learning_rate": 6.501973877387122e-05, "loss": 0.025, "step": 34620 }, { "grad_norm": 0.3864098787307739, "learning_rate": 6.500002368809098e-05, "loss": 0.0278, "step": 34630 }, { "grad_norm": 0.4287921190261841, "learning_rate": 6.498030603921694e-05, "loss": 0.0287, "step": 34640 }, { "grad_norm": 0.3509052097797394, "learning_rate": 6.496058583061832e-05, "loss": 0.0323, "step": 34650 }, { "grad_norm": 0.3399832546710968, "learning_rate": 6.494086306566475e-05, "loss": 0.0256, "step": 34660 }, { "grad_norm": 0.27953630685806274, "learning_rate": 6.492113774772632e-05, "loss": 0.0286, "step": 34670 }, { "grad_norm": 0.4603632688522339, "learning_rate": 6.490140988017354e-05, "loss": 0.0379, "step": 34680 }, { "grad_norm": 0.38779816031455994, "learning_rate": 6.488167946637736e-05, "loss": 0.0323, "step": 34690 }, { "grad_norm": 0.439049631357193, "learning_rate": 6.486194650970915e-05, "loss": 0.0356, "step": 34700 }, { "grad_norm": 0.3569556474685669, "learning_rate": 6.48422110135408e-05, "loss": 0.0331, "step": 34710 }, { "grad_norm": 0.3253948390483856, "learning_rate": 6.482247298124451e-05, "loss": 0.0426, "step": 34720 }, { "grad_norm": 0.3511698842048645, "learning_rate": 6.480273241619297e-05, "loss": 0.0259, "step": 34730 }, { "grad_norm": 0.3383777141571045, "learning_rate": 6.478298932175933e-05, "loss": 0.025, "step": 34740 }, { "grad_norm": 0.26915043592453003, "learning_rate": 6.476324370131712e-05, "loss": 0.0248, "step": 34750 }, { "grad_norm": 0.49985286593437195, "learning_rate": 6.474349555824036e-05, "loss": 0.0281, "step": 34760 }, { "grad_norm": 0.2576488256454468, "learning_rate": 6.472374489590342e-05, "loss": 0.0203, "step": 34770 }, { "grad_norm": 0.3572194576263428, "learning_rate": 6.470399171768118e-05, "loss": 0.0352, "step": 34780 }, { "grad_norm": 0.40250468254089355, "learning_rate": 6.468423602694891e-05, "loss": 0.0258, "step": 34790 }, { "grad_norm": 0.4541253447532654, "learning_rate": 6.466447782708232e-05, "loss": 0.0276, "step": 34800 }, { "grad_norm": 0.3809056878089905, "learning_rate": 6.464471712145754e-05, "loss": 0.0304, "step": 34810 }, { "grad_norm": 0.4963357150554657, "learning_rate": 6.462495391345114e-05, "loss": 0.0294, "step": 34820 }, { "grad_norm": 0.24317994713783264, "learning_rate": 6.46051882064401e-05, "loss": 0.0249, "step": 34830 }, { "grad_norm": 0.3876040279865265, "learning_rate": 6.458542000380186e-05, "loss": 0.0269, "step": 34840 }, { "grad_norm": 0.3281612694263458, "learning_rate": 6.456564930891424e-05, "loss": 0.0236, "step": 34850 }, { "grad_norm": 0.318280428647995, "learning_rate": 6.454587612515555e-05, "loss": 0.0285, "step": 34860 }, { "grad_norm": 0.38364216685295105, "learning_rate": 6.452610045590444e-05, "loss": 0.0317, "step": 34870 }, { "grad_norm": 0.36562567949295044, "learning_rate": 6.450632230454005e-05, "loss": 0.0266, "step": 34880 }, { "grad_norm": 0.4056698977947235, "learning_rate": 6.448654167444195e-05, "loss": 0.0372, "step": 34890 }, { "grad_norm": 0.3245074450969696, "learning_rate": 6.446675856899005e-05, "loss": 0.0226, "step": 34900 }, { "grad_norm": 0.3656175434589386, "learning_rate": 6.444697299156481e-05, "loss": 0.0286, "step": 34910 }, { "grad_norm": 0.5645647048950195, "learning_rate": 6.442718494554701e-05, "loss": 0.0292, "step": 34920 }, { "grad_norm": 0.6594809293746948, "learning_rate": 6.440739443431787e-05, "loss": 0.0246, "step": 34930 }, { "grad_norm": 0.28152263164520264, "learning_rate": 6.438760146125906e-05, "loss": 0.0273, "step": 34940 }, { "grad_norm": 0.4337044358253479, "learning_rate": 6.436780602975267e-05, "loss": 0.0495, "step": 34950 }, { "grad_norm": 0.48462000489234924, "learning_rate": 6.43480081431812e-05, "loss": 0.035, "step": 34960 }, { "grad_norm": 0.548877477645874, "learning_rate": 6.432820780492756e-05, "loss": 0.0407, "step": 34970 }, { "grad_norm": 0.30378258228302, "learning_rate": 6.430840501837506e-05, "loss": 0.0332, "step": 34980 }, { "grad_norm": 0.40251824259757996, "learning_rate": 6.428859978690748e-05, "loss": 0.0288, "step": 34990 }, { "grad_norm": 0.41157129406929016, "learning_rate": 6.426879211390901e-05, "loss": 0.0277, "step": 35000 }, { "grad_norm": 0.3681892454624176, "learning_rate": 6.424898200276422e-05, "loss": 0.0284, "step": 35010 }, { "grad_norm": 0.3600222170352936, "learning_rate": 6.42291694568581e-05, "loss": 0.0246, "step": 35020 }, { "grad_norm": 0.27877870202064514, "learning_rate": 6.42093544795761e-05, "loss": 0.0318, "step": 35030 }, { "grad_norm": 0.26270586252212524, "learning_rate": 6.418953707430403e-05, "loss": 0.0283, "step": 35040 }, { "grad_norm": 0.3466140329837799, "learning_rate": 6.416971724442819e-05, "loss": 0.0332, "step": 35050 }, { "grad_norm": 0.3560572862625122, "learning_rate": 6.414989499333519e-05, "loss": 0.0303, "step": 35060 }, { "grad_norm": 0.3479442000389099, "learning_rate": 6.413007032441214e-05, "loss": 0.0271, "step": 35070 }, { "grad_norm": 0.35743772983551025, "learning_rate": 6.411024324104653e-05, "loss": 0.0326, "step": 35080 }, { "grad_norm": 0.38187703490257263, "learning_rate": 6.409041374662628e-05, "loss": 0.0257, "step": 35090 }, { "grad_norm": 0.31046992540359497, "learning_rate": 6.407058184453967e-05, "loss": 0.0431, "step": 35100 }, { "grad_norm": 0.33515557646751404, "learning_rate": 6.405074753817548e-05, "loss": 0.0257, "step": 35110 }, { "grad_norm": 0.4139604866504669, "learning_rate": 6.40309108309228e-05, "loss": 0.027, "step": 35120 }, { "grad_norm": 0.3348778486251831, "learning_rate": 6.401107172617122e-05, "loss": 0.0254, "step": 35130 }, { "grad_norm": 0.47036635875701904, "learning_rate": 6.399123022731068e-05, "loss": 0.0344, "step": 35140 }, { "grad_norm": 0.36960938572883606, "learning_rate": 6.397138633773157e-05, "loss": 0.024, "step": 35150 }, { "grad_norm": 0.453667014837265, "learning_rate": 6.395154006082463e-05, "loss": 0.031, "step": 35160 }, { "grad_norm": 0.23431548476219177, "learning_rate": 6.393169139998109e-05, "loss": 0.0243, "step": 35170 }, { "grad_norm": 0.3026924431324005, "learning_rate": 6.39118403585925e-05, "loss": 0.0277, "step": 35180 }, { "grad_norm": 0.35151010751724243, "learning_rate": 6.38919869400509e-05, "loss": 0.0241, "step": 35190 }, { "grad_norm": 0.224432110786438, "learning_rate": 6.387213114774865e-05, "loss": 0.024, "step": 35200 }, { "grad_norm": 0.2578229606151581, "learning_rate": 6.385227298507863e-05, "loss": 0.0442, "step": 35210 }, { "grad_norm": 0.3710765540599823, "learning_rate": 6.3832412455434e-05, "loss": 0.0197, "step": 35220 }, { "grad_norm": 0.17656779289245605, "learning_rate": 6.381254956220841e-05, "loss": 0.026, "step": 35230 }, { "grad_norm": 0.24947921931743622, "learning_rate": 6.379268430879586e-05, "loss": 0.0248, "step": 35240 }, { "grad_norm": 0.3126426637172699, "learning_rate": 6.37728166985908e-05, "loss": 0.0227, "step": 35250 }, { "grad_norm": 0.33841031789779663, "learning_rate": 6.375294673498804e-05, "loss": 0.0219, "step": 35260 }, { "grad_norm": 0.3945915102958679, "learning_rate": 6.373307442138284e-05, "loss": 0.0258, "step": 35270 }, { "grad_norm": 0.33751827478408813, "learning_rate": 6.371319976117081e-05, "loss": 0.0331, "step": 35280 }, { "grad_norm": 0.31293216347694397, "learning_rate": 6.3693322757748e-05, "loss": 0.0272, "step": 35290 }, { "grad_norm": 0.3273959159851074, "learning_rate": 6.367344341451086e-05, "loss": 0.0338, "step": 35300 }, { "grad_norm": 0.3552035391330719, "learning_rate": 6.36535617348562e-05, "loss": 0.0294, "step": 35310 }, { "grad_norm": 0.257950097322464, "learning_rate": 6.363367772218128e-05, "loss": 0.0244, "step": 35320 }, { "grad_norm": 0.5122740268707275, "learning_rate": 6.36137913798837e-05, "loss": 0.03, "step": 35330 }, { "grad_norm": 0.32079020142555237, "learning_rate": 6.359390271136151e-05, "loss": 0.0221, "step": 35340 }, { "grad_norm": 0.28706154227256775, "learning_rate": 6.357401172001314e-05, "loss": 0.027, "step": 35350 }, { "grad_norm": 0.21675889194011688, "learning_rate": 6.355411840923742e-05, "loss": 0.0225, "step": 35360 }, { "grad_norm": 0.441202849149704, "learning_rate": 6.353422278243358e-05, "loss": 0.0244, "step": 35370 }, { "grad_norm": 0.2899806797504425, "learning_rate": 6.351432484300121e-05, "loss": 0.0267, "step": 35380 }, { "grad_norm": 0.362214595079422, "learning_rate": 6.349442459434036e-05, "loss": 0.0302, "step": 35390 }, { "grad_norm": 0.29193368554115295, "learning_rate": 6.34745220398514e-05, "loss": 0.0327, "step": 35400 }, { "grad_norm": 0.3439536392688751, "learning_rate": 6.345461718293518e-05, "loss": 0.0272, "step": 35410 }, { "grad_norm": 0.26206380128860474, "learning_rate": 6.343471002699286e-05, "loss": 0.0304, "step": 35420 }, { "grad_norm": 0.39500191807746887, "learning_rate": 6.341480057542602e-05, "loss": 0.0339, "step": 35430 }, { "grad_norm": 0.36163824796676636, "learning_rate": 6.339488883163667e-05, "loss": 0.0268, "step": 35440 }, { "grad_norm": 0.2215055525302887, "learning_rate": 6.337497479902716e-05, "loss": 0.0232, "step": 35450 }, { "grad_norm": 0.32250165939331055, "learning_rate": 6.335505848100027e-05, "loss": 0.0248, "step": 35460 }, { "grad_norm": 0.467913955450058, "learning_rate": 6.333513988095915e-05, "loss": 0.0205, "step": 35470 }, { "grad_norm": 0.20813816785812378, "learning_rate": 6.331521900230735e-05, "loss": 0.0221, "step": 35480 }, { "grad_norm": 0.25169962644577026, "learning_rate": 6.329529584844878e-05, "loss": 0.0211, "step": 35490 }, { "grad_norm": 0.3904198408126831, "learning_rate": 6.327537042278777e-05, "loss": 0.0297, "step": 35500 }, { "grad_norm": 0.35288986563682556, "learning_rate": 6.325544272872905e-05, "loss": 0.0253, "step": 35510 }, { "grad_norm": 0.34000012278556824, "learning_rate": 6.323551276967771e-05, "loss": 0.0327, "step": 35520 }, { "grad_norm": 0.3243982195854187, "learning_rate": 6.321558054903922e-05, "loss": 0.0234, "step": 35530 }, { "grad_norm": 0.30828893184661865, "learning_rate": 6.319564607021947e-05, "loss": 0.0287, "step": 35540 }, { "grad_norm": 0.26038867235183716, "learning_rate": 6.31757093366247e-05, "loss": 0.0216, "step": 35550 }, { "grad_norm": 0.2826596796512604, "learning_rate": 6.315577035166154e-05, "loss": 0.0253, "step": 35560 }, { "grad_norm": 0.39104846119880676, "learning_rate": 6.313582911873708e-05, "loss": 0.0234, "step": 35570 }, { "grad_norm": 0.4188324511051178, "learning_rate": 6.311588564125865e-05, "loss": 0.0222, "step": 35580 }, { "grad_norm": 0.5268115997314453, "learning_rate": 6.30959399226341e-05, "loss": 0.0322, "step": 35590 }, { "grad_norm": 0.3305550813674927, "learning_rate": 6.30759919662716e-05, "loss": 0.03, "step": 35600 }, { "grad_norm": 0.2746639549732208, "learning_rate": 6.30560417755797e-05, "loss": 0.0235, "step": 35610 }, { "grad_norm": 0.16305208206176758, "learning_rate": 6.303608935396735e-05, "loss": 0.0187, "step": 35620 }, { "grad_norm": 0.27492472529411316, "learning_rate": 6.301613470484386e-05, "loss": 0.0244, "step": 35630 }, { "grad_norm": 0.33177870512008667, "learning_rate": 6.299617783161893e-05, "loss": 0.0251, "step": 35640 }, { "grad_norm": 0.2779829502105713, "learning_rate": 6.297621873770266e-05, "loss": 0.0264, "step": 35650 }, { "grad_norm": 0.2989441752433777, "learning_rate": 6.29562574265055e-05, "loss": 0.0267, "step": 35660 }, { "grad_norm": 0.43246424198150635, "learning_rate": 6.293629390143834e-05, "loss": 0.0248, "step": 35670 }, { "grad_norm": 0.35892391204833984, "learning_rate": 6.291632816591232e-05, "loss": 0.0225, "step": 35680 }, { "grad_norm": 0.2286178469657898, "learning_rate": 6.28963602233391e-05, "loss": 0.0195, "step": 35690 }, { "grad_norm": 0.36206814646720886, "learning_rate": 6.287639007713062e-05, "loss": 0.029, "step": 35700 }, { "grad_norm": 0.2978808283805847, "learning_rate": 6.285641773069926e-05, "loss": 0.025, "step": 35710 }, { "grad_norm": 0.2497752159833908, "learning_rate": 6.283644318745773e-05, "loss": 0.0187, "step": 35720 }, { "grad_norm": 0.31654995679855347, "learning_rate": 6.281646645081912e-05, "loss": 0.0249, "step": 35730 }, { "grad_norm": 0.3311185836791992, "learning_rate": 6.279648752419693e-05, "loss": 0.025, "step": 35740 }, { "grad_norm": 0.3807659447193146, "learning_rate": 6.2776506411005e-05, "loss": 0.0254, "step": 35750 }, { "grad_norm": 0.4190659523010254, "learning_rate": 6.275652311465758e-05, "loss": 0.0231, "step": 35760 }, { "grad_norm": 0.37657442688941956, "learning_rate": 6.273653763856926e-05, "loss": 0.0219, "step": 35770 }, { "grad_norm": 0.3357340693473816, "learning_rate": 6.271654998615501e-05, "loss": 0.0373, "step": 35780 }, { "grad_norm": 0.2802472412586212, "learning_rate": 6.269656016083013e-05, "loss": 0.0227, "step": 35790 }, { "grad_norm": 0.3748089671134949, "learning_rate": 6.267656816601038e-05, "loss": 0.0335, "step": 35800 }, { "grad_norm": 0.3800283968448639, "learning_rate": 6.265657400511185e-05, "loss": 0.0275, "step": 35810 }, { "grad_norm": 0.36452293395996094, "learning_rate": 6.263657768155098e-05, "loss": 0.0279, "step": 35820 }, { "grad_norm": 0.3030935823917389, "learning_rate": 6.261657919874457e-05, "loss": 0.0253, "step": 35830 }, { "grad_norm": 0.4274923503398895, "learning_rate": 6.259657856010986e-05, "loss": 0.0336, "step": 35840 }, { "grad_norm": 0.30974432826042175, "learning_rate": 6.257657576906439e-05, "loss": 0.0226, "step": 35850 }, { "grad_norm": 0.5442358255386353, "learning_rate": 6.255657082902609e-05, "loss": 0.0247, "step": 35860 }, { "grad_norm": 0.3276223838329315, "learning_rate": 6.253656374341325e-05, "loss": 0.0259, "step": 35870 }, { "grad_norm": 0.34904730319976807, "learning_rate": 6.251655451564457e-05, "loss": 0.0277, "step": 35880 }, { "grad_norm": 0.30127546191215515, "learning_rate": 6.249654314913902e-05, "loss": 0.0207, "step": 35890 }, { "grad_norm": 0.2931177318096161, "learning_rate": 6.247652964731604e-05, "loss": 0.0288, "step": 35900 }, { "grad_norm": 0.24212108552455902, "learning_rate": 6.245651401359537e-05, "loss": 0.0482, "step": 35910 }, { "grad_norm": 0.2930770814418793, "learning_rate": 6.243649625139715e-05, "loss": 0.0267, "step": 35920 }, { "grad_norm": 0.27419009804725647, "learning_rate": 6.241647636414185e-05, "loss": 0.0252, "step": 35930 }, { "grad_norm": 0.6129536032676697, "learning_rate": 6.239645435525034e-05, "loss": 0.0319, "step": 35940 }, { "grad_norm": 0.2344682514667511, "learning_rate": 6.237643022814381e-05, "loss": 0.0237, "step": 35950 }, { "grad_norm": 0.39909034967422485, "learning_rate": 6.235640398624386e-05, "loss": 0.0235, "step": 35960 }, { "grad_norm": 0.3333500325679779, "learning_rate": 6.233637563297243e-05, "loss": 0.0289, "step": 35970 }, { "grad_norm": 0.418817400932312, "learning_rate": 6.23163451717518e-05, "loss": 0.0303, "step": 35980 }, { "grad_norm": 0.3718324899673462, "learning_rate": 6.229631260600463e-05, "loss": 0.0251, "step": 35990 }, { "grad_norm": 0.3407174348831177, "learning_rate": 6.227627793915392e-05, "loss": 0.0318, "step": 36000 }, { "grad_norm": 0.3477597236633301, "learning_rate": 6.225624117462309e-05, "loss": 0.0218, "step": 36010 }, { "grad_norm": 0.41588735580444336, "learning_rate": 6.223620231583586e-05, "loss": 0.0229, "step": 36020 }, { "grad_norm": 0.29528704285621643, "learning_rate": 6.221616136621629e-05, "loss": 0.0243, "step": 36030 }, { "grad_norm": 0.3113083839416504, "learning_rate": 6.219611832918887e-05, "loss": 0.0248, "step": 36040 }, { "grad_norm": 0.231722891330719, "learning_rate": 6.217607320817838e-05, "loss": 0.021, "step": 36050 }, { "grad_norm": 0.255503386259079, "learning_rate": 6.215602600661001e-05, "loss": 0.0223, "step": 36060 }, { "grad_norm": 0.4247032403945923, "learning_rate": 6.213597672790925e-05, "loss": 0.0282, "step": 36070 }, { "grad_norm": 0.3156534433364868, "learning_rate": 6.2115925375502e-05, "loss": 0.0239, "step": 36080 }, { "grad_norm": 0.5425737500190735, "learning_rate": 6.209587195281447e-05, "loss": 0.0304, "step": 36090 }, { "grad_norm": 0.37397024035453796, "learning_rate": 6.207581646327324e-05, "loss": 0.0286, "step": 36100 }, { "grad_norm": 0.19213716685771942, "learning_rate": 6.205575891030526e-05, "loss": 0.018, "step": 36110 }, { "grad_norm": 0.23696719110012054, "learning_rate": 6.203569929733781e-05, "loss": 0.024, "step": 36120 }, { "grad_norm": 0.2950247526168823, "learning_rate": 6.201563762779852e-05, "loss": 0.023, "step": 36130 }, { "grad_norm": 0.19082234799861908, "learning_rate": 6.199557390511538e-05, "loss": 0.022, "step": 36140 }, { "grad_norm": 0.2851802408695221, "learning_rate": 6.197550813271675e-05, "loss": 0.0273, "step": 36150 }, { "grad_norm": 0.2959071397781372, "learning_rate": 6.195544031403131e-05, "loss": 0.032, "step": 36160 }, { "grad_norm": 0.3973654508590698, "learning_rate": 6.19353704524881e-05, "loss": 0.0239, "step": 36170 }, { "grad_norm": 0.19388826191425323, "learning_rate": 6.191529855151652e-05, "loss": 0.0212, "step": 36180 }, { "grad_norm": 0.36229315400123596, "learning_rate": 6.189522461454629e-05, "loss": 0.024, "step": 36190 }, { "grad_norm": 0.2957962155342102, "learning_rate": 6.187514864500752e-05, "loss": 0.0267, "step": 36200 }, { "grad_norm": 0.4853380620479584, "learning_rate": 6.185507064633062e-05, "loss": 0.0272, "step": 36210 }, { "grad_norm": 0.4685152471065521, "learning_rate": 6.18349906219464e-05, "loss": 0.0259, "step": 36220 }, { "grad_norm": 0.5084390044212341, "learning_rate": 6.181490857528596e-05, "loss": 0.0262, "step": 36230 }, { "grad_norm": 0.23267078399658203, "learning_rate": 6.179482450978077e-05, "loss": 0.0245, "step": 36240 }, { "grad_norm": 0.3812970817089081, "learning_rate": 6.177473842886269e-05, "loss": 0.0285, "step": 36250 }, { "grad_norm": 0.39053481817245483, "learning_rate": 6.175465033596382e-05, "loss": 0.0313, "step": 36260 }, { "grad_norm": 0.40597212314605713, "learning_rate": 6.173456023451671e-05, "loss": 0.027, "step": 36270 }, { "grad_norm": 0.4273182451725006, "learning_rate": 6.171446812795422e-05, "loss": 0.0297, "step": 36280 }, { "grad_norm": 0.27366402745246887, "learning_rate": 6.169437401970949e-05, "loss": 0.0378, "step": 36290 }, { "grad_norm": 0.19202820956707, "learning_rate": 6.16742779132161e-05, "loss": 0.0196, "step": 36300 }, { "grad_norm": 0.31078943610191345, "learning_rate": 6.165417981190789e-05, "loss": 0.0246, "step": 36310 }, { "grad_norm": 0.5683124661445618, "learning_rate": 6.16340797192191e-05, "loss": 0.0224, "step": 36320 }, { "grad_norm": 0.2598726451396942, "learning_rate": 6.161397763858427e-05, "loss": 0.0223, "step": 36330 }, { "grad_norm": 0.28237640857696533, "learning_rate": 6.159387357343834e-05, "loss": 0.0198, "step": 36340 }, { "grad_norm": 0.3297339081764221, "learning_rate": 6.157376752721648e-05, "loss": 0.0246, "step": 36350 }, { "grad_norm": 0.2643367052078247, "learning_rate": 6.155365950335428e-05, "loss": 0.0272, "step": 36360 }, { "grad_norm": 0.43473029136657715, "learning_rate": 6.153354950528768e-05, "loss": 0.0232, "step": 36370 }, { "grad_norm": 0.35762375593185425, "learning_rate": 6.151343753645293e-05, "loss": 0.0272, "step": 36380 }, { "grad_norm": 0.33919909596443176, "learning_rate": 6.149332360028657e-05, "loss": 0.0335, "step": 36390 }, { "grad_norm": 0.2944895923137665, "learning_rate": 6.147320770022555e-05, "loss": 0.0217, "step": 36400 }, { "grad_norm": 0.2787952721118927, "learning_rate": 6.145308983970715e-05, "loss": 0.0281, "step": 36410 }, { "grad_norm": 0.20488637685775757, "learning_rate": 6.143297002216892e-05, "loss": 0.0239, "step": 36420 }, { "grad_norm": 0.23931413888931274, "learning_rate": 6.141284825104882e-05, "loss": 0.0231, "step": 36430 }, { "grad_norm": 0.3451227843761444, "learning_rate": 6.13927245297851e-05, "loss": 0.0323, "step": 36440 }, { "grad_norm": 0.3622733950614929, "learning_rate": 6.137259886181633e-05, "loss": 0.0329, "step": 36450 }, { "grad_norm": 0.40176814794540405, "learning_rate": 6.135247125058145e-05, "loss": 0.0333, "step": 36460 }, { "grad_norm": 0.4912019371986389, "learning_rate": 6.133234169951974e-05, "loss": 0.0206, "step": 36470 }, { "grad_norm": 0.26191529631614685, "learning_rate": 6.131221021207078e-05, "loss": 0.0304, "step": 36480 }, { "grad_norm": 0.2695438861846924, "learning_rate": 6.129207679167448e-05, "loss": 0.0192, "step": 36490 }, { "grad_norm": 0.22827661037445068, "learning_rate": 6.127194144177109e-05, "loss": 0.0187, "step": 36500 }, { "grad_norm": 0.3652871549129486, "learning_rate": 6.125180416580118e-05, "loss": 0.019, "step": 36510 }, { "grad_norm": 0.21530631184577942, "learning_rate": 6.123166496720571e-05, "loss": 0.0208, "step": 36520 }, { "grad_norm": 0.23142993450164795, "learning_rate": 6.121152384942588e-05, "loss": 0.0241, "step": 36530 }, { "grad_norm": 0.5191365480422974, "learning_rate": 6.119138081590324e-05, "loss": 0.0269, "step": 36540 }, { "grad_norm": 0.27076300978660583, "learning_rate": 6.117123587007971e-05, "loss": 0.0199, "step": 36550 }, { "grad_norm": 0.26432669162750244, "learning_rate": 6.11510890153975e-05, "loss": 0.0294, "step": 36560 }, { "grad_norm": 0.2602804899215698, "learning_rate": 6.113094025529916e-05, "loss": 0.0228, "step": 36570 }, { "grad_norm": 0.31779760122299194, "learning_rate": 6.111078959322757e-05, "loss": 0.0261, "step": 36580 }, { "grad_norm": 0.40462300181388855, "learning_rate": 6.109063703262592e-05, "loss": 0.0311, "step": 36590 }, { "grad_norm": 0.3988780677318573, "learning_rate": 6.107048257693772e-05, "loss": 0.029, "step": 36600 }, { "grad_norm": 0.2509232461452484, "learning_rate": 6.105032622960683e-05, "loss": 0.0243, "step": 36610 }, { "grad_norm": 0.2964089810848236, "learning_rate": 6.103016799407743e-05, "loss": 0.02, "step": 36620 }, { "grad_norm": 0.31579461693763733, "learning_rate": 6.1010007873793984e-05, "loss": 0.0278, "step": 36630 }, { "grad_norm": 0.3478187322616577, "learning_rate": 6.098984587220131e-05, "loss": 0.0238, "step": 36640 }, { "grad_norm": 0.32869845628738403, "learning_rate": 6.096968199274456e-05, "loss": 0.0262, "step": 36650 }, { "grad_norm": 0.3511703908443451, "learning_rate": 6.0949516238869166e-05, "loss": 0.027, "step": 36660 }, { "grad_norm": 0.3705185651779175, "learning_rate": 6.092934861402092e-05, "loss": 0.03, "step": 36670 }, { "grad_norm": 0.34984663128852844, "learning_rate": 6.0909179121645924e-05, "loss": 0.024, "step": 36680 }, { "grad_norm": 0.3829476535320282, "learning_rate": 6.0889007765190576e-05, "loss": 0.0278, "step": 36690 }, { "grad_norm": 0.2513662576675415, "learning_rate": 6.086883454810162e-05, "loss": 0.0205, "step": 36700 }, { "grad_norm": 0.29126638174057007, "learning_rate": 6.0848659473826084e-05, "loss": 0.0208, "step": 36710 }, { "grad_norm": 0.37374603748321533, "learning_rate": 6.082848254581138e-05, "loss": 0.0282, "step": 36720 }, { "grad_norm": 0.3980967104434967, "learning_rate": 6.080830376750517e-05, "loss": 0.0283, "step": 36730 }, { "grad_norm": 0.26289811730384827, "learning_rate": 6.0788123142355445e-05, "loss": 0.0218, "step": 36740 }, { "grad_norm": 0.2999984920024872, "learning_rate": 6.076794067381052e-05, "loss": 0.0261, "step": 36750 }, { "grad_norm": 0.1971672922372818, "learning_rate": 6.074775636531905e-05, "loss": 0.0231, "step": 36760 }, { "grad_norm": 0.4190150797367096, "learning_rate": 6.072757022032997e-05, "loss": 0.0217, "step": 36770 }, { "grad_norm": 0.2811511158943176, "learning_rate": 6.070738224229253e-05, "loss": 0.0312, "step": 36780 }, { "grad_norm": 0.26763126254081726, "learning_rate": 6.0687192434656314e-05, "loss": 0.0299, "step": 36790 }, { "grad_norm": 0.5314334630966187, "learning_rate": 6.066700080087121e-05, "loss": 0.0261, "step": 36800 }, { "grad_norm": 0.242157980799675, "learning_rate": 6.0646807344387424e-05, "loss": 0.0222, "step": 36810 }, { "grad_norm": 0.3706907331943512, "learning_rate": 6.062661206865543e-05, "loss": 0.0315, "step": 36820 }, { "grad_norm": 0.20830750465393066, "learning_rate": 6.06064149771261e-05, "loss": 0.0229, "step": 36830 }, { "grad_norm": 0.3340285122394562, "learning_rate": 6.058621607325051e-05, "loss": 0.0239, "step": 36840 }, { "grad_norm": 0.35705670714378357, "learning_rate": 6.056601536048014e-05, "loss": 0.0281, "step": 36850 }, { "grad_norm": 0.27511611580848694, "learning_rate": 6.0545812842266725e-05, "loss": 0.0315, "step": 36860 }, { "grad_norm": 0.2844195067882538, "learning_rate": 6.052560852206232e-05, "loss": 0.0294, "step": 36870 }, { "grad_norm": 0.29982754588127136, "learning_rate": 6.05054024033193e-05, "loss": 0.0338, "step": 36880 }, { "grad_norm": 0.31757596135139465, "learning_rate": 6.048519448949032e-05, "loss": 0.0254, "step": 36890 }, { "grad_norm": 0.29139482975006104, "learning_rate": 6.046498478402839e-05, "loss": 0.0304, "step": 36900 }, { "grad_norm": 0.32586368918418884, "learning_rate": 6.044477329038677e-05, "loss": 0.0274, "step": 36910 }, { "grad_norm": 0.31845995783805847, "learning_rate": 6.042456001201906e-05, "loss": 0.0219, "step": 36920 }, { "grad_norm": 0.21939688920974731, "learning_rate": 6.040434495237917e-05, "loss": 0.0255, "step": 36930 }, { "grad_norm": 0.5262258648872375, "learning_rate": 6.0384128114921256e-05, "loss": 0.0294, "step": 36940 }, { "grad_norm": 0.3356195092201233, "learning_rate": 6.036390950309987e-05, "loss": 0.0185, "step": 36950 }, { "grad_norm": 0.3773374855518341, "learning_rate": 6.0343689120369805e-05, "loss": 0.0265, "step": 36960 }, { "grad_norm": 0.2912042438983917, "learning_rate": 6.032346697018616e-05, "loss": 0.0215, "step": 36970 }, { "grad_norm": 0.2984762191772461, "learning_rate": 6.0303243056004375e-05, "loss": 0.0339, "step": 36980 }, { "grad_norm": 0.23008808493614197, "learning_rate": 6.0283017381280136e-05, "loss": 0.0226, "step": 36990 }, { "grad_norm": 0.4372289776802063, "learning_rate": 6.026278994946948e-05, "loss": 0.0222, "step": 37000 }, { "grad_norm": 0.2837933897972107, "learning_rate": 6.02425607640287e-05, "loss": 0.0258, "step": 37010 }, { "grad_norm": 0.3809496760368347, "learning_rate": 6.022232982841441e-05, "loss": 0.0266, "step": 37020 }, { "grad_norm": 0.33108028769493103, "learning_rate": 6.020209714608355e-05, "loss": 0.017, "step": 37030 }, { "grad_norm": 0.332118421792984, "learning_rate": 6.018186272049331e-05, "loss": 0.0263, "step": 37040 }, { "grad_norm": 0.33946284651756287, "learning_rate": 6.01616265551012e-05, "loss": 0.033, "step": 37050 }, { "grad_norm": 0.3900570869445801, "learning_rate": 6.014138865336503e-05, "loss": 0.0252, "step": 37060 }, { "grad_norm": 0.2386033684015274, "learning_rate": 6.0121149018742905e-05, "loss": 0.032, "step": 37070 }, { "grad_norm": 0.32510238885879517, "learning_rate": 6.010090765469325e-05, "loss": 0.0265, "step": 37080 }, { "grad_norm": 0.36636900901794434, "learning_rate": 6.008066456467473e-05, "loss": 0.0225, "step": 37090 }, { "grad_norm": 0.2636350095272064, "learning_rate": 6.0060419752146335e-05, "loss": 0.025, "step": 37100 }, { "grad_norm": 0.3016299307346344, "learning_rate": 6.0040173220567353e-05, "loss": 0.0237, "step": 37110 }, { "grad_norm": 0.26417654752731323, "learning_rate": 6.001992497339737e-05, "loss": 0.0248, "step": 37120 }, { "grad_norm": 0.29285675287246704, "learning_rate": 5.999967501409626e-05, "loss": 0.0244, "step": 37130 }, { "grad_norm": 0.22417959570884705, "learning_rate": 5.997942334612418e-05, "loss": 0.0202, "step": 37140 }, { "grad_norm": 0.3478032350540161, "learning_rate": 5.995916997294158e-05, "loss": 0.0232, "step": 37150 }, { "grad_norm": 0.4246840178966522, "learning_rate": 5.9938914898009214e-05, "loss": 0.0259, "step": 37160 }, { "grad_norm": 0.3043409585952759, "learning_rate": 5.991865812478813e-05, "loss": 0.0185, "step": 37170 }, { "grad_norm": 0.4290139675140381, "learning_rate": 5.989839965673964e-05, "loss": 0.0361, "step": 37180 }, { "grad_norm": 0.18283727765083313, "learning_rate": 5.987813949732539e-05, "loss": 0.0261, "step": 37190 }, { "grad_norm": 0.25287020206451416, "learning_rate": 5.9857877650007255e-05, "loss": 0.0196, "step": 37200 }, { "grad_norm": 0.3725214898586273, "learning_rate": 5.983761411824744e-05, "loss": 0.0234, "step": 37210 }, { "grad_norm": 0.32438284158706665, "learning_rate": 5.981734890550844e-05, "loss": 0.0296, "step": 37220 }, { "grad_norm": 0.31092917919158936, "learning_rate": 5.979708201525301e-05, "loss": 0.0319, "step": 37230 }, { "grad_norm": 0.4753456115722656, "learning_rate": 5.977681345094422e-05, "loss": 0.0277, "step": 37240 }, { "grad_norm": 0.2869553565979004, "learning_rate": 5.97565432160454e-05, "loss": 0.0293, "step": 37250 }, { "grad_norm": 0.34261882305145264, "learning_rate": 5.9736271314020186e-05, "loss": 0.0293, "step": 37260 }, { "grad_norm": 0.28289157152175903, "learning_rate": 5.971599774833251e-05, "loss": 0.0257, "step": 37270 }, { "grad_norm": 0.46545588970184326, "learning_rate": 5.9695722522446525e-05, "loss": 0.032, "step": 37280 }, { "grad_norm": 1.1166598796844482, "learning_rate": 5.9675445639826765e-05, "loss": 0.0334, "step": 37290 }, { "grad_norm": 0.29583004117012024, "learning_rate": 5.965516710393796e-05, "loss": 0.0277, "step": 37300 }, { "grad_norm": 0.47735080122947693, "learning_rate": 5.963488691824516e-05, "loss": 0.0348, "step": 37310 }, { "grad_norm": 0.35471218824386597, "learning_rate": 5.96146050862137e-05, "loss": 0.0293, "step": 37320 }, { "grad_norm": 0.33873680233955383, "learning_rate": 5.959432161130919e-05, "loss": 0.0247, "step": 37330 }, { "grad_norm": 0.2662562429904938, "learning_rate": 5.9574036496997545e-05, "loss": 0.0337, "step": 37340 }, { "grad_norm": 0.2821158468723297, "learning_rate": 5.955374974674488e-05, "loss": 0.0274, "step": 37350 }, { "grad_norm": 0.3602330982685089, "learning_rate": 5.9533461364017696e-05, "loss": 0.0241, "step": 37360 }, { "grad_norm": 0.3213043808937073, "learning_rate": 5.9513171352282716e-05, "loss": 0.0249, "step": 37370 }, { "grad_norm": 0.39123231172561646, "learning_rate": 5.949287971500692e-05, "loss": 0.0265, "step": 37380 }, { "grad_norm": 0.2431226670742035, "learning_rate": 5.947258645565762e-05, "loss": 0.0201, "step": 37390 }, { "grad_norm": 0.19719642400741577, "learning_rate": 5.945229157770237e-05, "loss": 0.0193, "step": 37400 }, { "grad_norm": 0.4703332483768463, "learning_rate": 5.9431995084609006e-05, "loss": 0.0253, "step": 37410 }, { "grad_norm": 0.31751498579978943, "learning_rate": 5.941169697984564e-05, "loss": 0.0242, "step": 37420 }, { "grad_norm": 0.3070685863494873, "learning_rate": 5.9391397266880675e-05, "loss": 0.0248, "step": 37430 }, { "grad_norm": 0.42734789848327637, "learning_rate": 5.937109594918279e-05, "loss": 0.0336, "step": 37440 }, { "grad_norm": 0.31681549549102783, "learning_rate": 5.9350793030220884e-05, "loss": 0.0331, "step": 37450 }, { "grad_norm": 0.2475919872522354, "learning_rate": 5.933048851346421e-05, "loss": 0.019, "step": 37460 }, { "grad_norm": 0.3103955388069153, "learning_rate": 5.931018240238224e-05, "loss": 0.0386, "step": 37470 }, { "grad_norm": 0.48353445529937744, "learning_rate": 5.928987470044471e-05, "loss": 0.0214, "step": 37480 }, { "grad_norm": 0.4679102301597595, "learning_rate": 5.9269565411121695e-05, "loss": 0.0259, "step": 37490 }, { "grad_norm": 0.30406713485717773, "learning_rate": 5.924925453788347e-05, "loss": 0.0255, "step": 37500 }, { "grad_norm": 0.21736811101436615, "learning_rate": 5.92289420842006e-05, "loss": 0.0289, "step": 37510 }, { "grad_norm": 0.4829616844654083, "learning_rate": 5.9208628053543945e-05, "loss": 0.0296, "step": 37520 }, { "grad_norm": 0.5372331142425537, "learning_rate": 5.918831244938462e-05, "loss": 0.0265, "step": 37530 }, { "grad_norm": 0.4336627721786499, "learning_rate": 5.9167995275194e-05, "loss": 0.0283, "step": 37540 }, { "grad_norm": 0.34976959228515625, "learning_rate": 5.914767653444373e-05, "loss": 0.0242, "step": 37550 }, { "grad_norm": 0.5046740174293518, "learning_rate": 5.912735623060572e-05, "loss": 0.0299, "step": 37560 }, { "grad_norm": 0.22435975074768066, "learning_rate": 5.910703436715217e-05, "loss": 0.021, "step": 37570 }, { "grad_norm": 0.1846349835395813, "learning_rate": 5.908671094755552e-05, "loss": 0.0222, "step": 37580 }, { "grad_norm": 0.31409895420074463, "learning_rate": 5.906638597528851e-05, "loss": 0.0257, "step": 37590 }, { "grad_norm": 0.32021114230155945, "learning_rate": 5.9046059453824076e-05, "loss": 0.0233, "step": 37600 }, { "grad_norm": 0.3109167218208313, "learning_rate": 5.9025731386635505e-05, "loss": 0.0202, "step": 37610 }, { "grad_norm": 0.43930599093437195, "learning_rate": 5.900540177719629e-05, "loss": 0.0223, "step": 37620 }, { "grad_norm": 0.408480167388916, "learning_rate": 5.898507062898021e-05, "loss": 0.027, "step": 37630 }, { "grad_norm": 0.2769339382648468, "learning_rate": 5.8964737945461316e-05, "loss": 0.0251, "step": 37640 }, { "grad_norm": 0.175127774477005, "learning_rate": 5.8944403730113885e-05, "loss": 0.0245, "step": 37650 }, { "grad_norm": 0.45148470997810364, "learning_rate": 5.892406798641248e-05, "loss": 0.0248, "step": 37660 }, { "grad_norm": 0.34358111023902893, "learning_rate": 5.890373071783193e-05, "loss": 0.0244, "step": 37670 }, { "grad_norm": 0.32534825801849365, "learning_rate": 5.888339192784732e-05, "loss": 0.0226, "step": 37680 }, { "grad_norm": 0.22647863626480103, "learning_rate": 5.8863051619934003e-05, "loss": 0.0244, "step": 37690 }, { "grad_norm": 0.33923932909965515, "learning_rate": 5.8842709797567554e-05, "loss": 0.0236, "step": 37700 }, { "grad_norm": 0.28204232454299927, "learning_rate": 5.8822366464223855e-05, "loss": 0.0189, "step": 37710 }, { "grad_norm": 0.2611443102359772, "learning_rate": 5.880202162337901e-05, "loss": 0.0285, "step": 37720 }, { "grad_norm": 0.33878347277641296, "learning_rate": 5.8781675278509405e-05, "loss": 0.0181, "step": 37730 }, { "grad_norm": 0.8914328813552856, "learning_rate": 5.8761327433091696e-05, "loss": 0.0206, "step": 37740 }, { "grad_norm": 0.458427369594574, "learning_rate": 5.874097809060275e-05, "loss": 0.0314, "step": 37750 }, { "grad_norm": 0.3996213376522064, "learning_rate": 5.87206272545197e-05, "loss": 0.0392, "step": 37760 }, { "grad_norm": 0.22039547562599182, "learning_rate": 5.8700274928319955e-05, "loss": 0.0255, "step": 37770 }, { "grad_norm": 0.37720826268196106, "learning_rate": 5.867992111548118e-05, "loss": 0.025, "step": 37780 }, { "grad_norm": 0.43655824661254883, "learning_rate": 5.865956581948131e-05, "loss": 0.024, "step": 37790 }, { "grad_norm": 0.3385656476020813, "learning_rate": 5.863920904379845e-05, "loss": 0.0228, "step": 37800 }, { "grad_norm": 0.3192431628704071, "learning_rate": 5.8618850791911064e-05, "loss": 0.0212, "step": 37810 }, { "grad_norm": 0.20977957546710968, "learning_rate": 5.859849106729779e-05, "loss": 0.0225, "step": 37820 }, { "grad_norm": 0.2615043818950653, "learning_rate": 5.857812987343758e-05, "loss": 0.025, "step": 37830 }, { "grad_norm": 0.4666856527328491, "learning_rate": 5.855776721380957e-05, "loss": 0.0279, "step": 37840 }, { "grad_norm": 0.4875471293926239, "learning_rate": 5.8537403091893217e-05, "loss": 0.0255, "step": 37850 }, { "grad_norm": 0.35808488726615906, "learning_rate": 5.851703751116816e-05, "loss": 0.0253, "step": 37860 }, { "grad_norm": 0.3614524304866791, "learning_rate": 5.8496670475114336e-05, "loss": 0.0217, "step": 37870 }, { "grad_norm": 0.3940526247024536, "learning_rate": 5.84763019872119e-05, "loss": 0.0303, "step": 37880 }, { "grad_norm": 0.23039118945598602, "learning_rate": 5.845593205094131e-05, "loss": 0.02, "step": 37890 }, { "grad_norm": 0.3487291634082794, "learning_rate": 5.843556066978318e-05, "loss": 0.0197, "step": 37900 }, { "grad_norm": 0.24778780341148376, "learning_rate": 5.8415187847218455e-05, "loss": 0.0249, "step": 37910 }, { "grad_norm": 0.4098760187625885, "learning_rate": 5.839481358672827e-05, "loss": 0.0233, "step": 37920 }, { "grad_norm": 0.4945228099822998, "learning_rate": 5.837443789179407e-05, "loss": 0.0239, "step": 37930 }, { "grad_norm": 0.31237900257110596, "learning_rate": 5.8354060765897445e-05, "loss": 0.0281, "step": 37940 }, { "grad_norm": 0.277062326669693, "learning_rate": 5.8333682212520334e-05, "loss": 0.0267, "step": 37950 }, { "grad_norm": 0.4835718274116516, "learning_rate": 5.831330223514486e-05, "loss": 0.0281, "step": 37960 }, { "grad_norm": 0.29736068844795227, "learning_rate": 5.8292920837253396e-05, "loss": 0.0272, "step": 37970 }, { "grad_norm": 0.22127844393253326, "learning_rate": 5.827253802232857e-05, "loss": 0.0188, "step": 37980 }, { "grad_norm": 0.47109416127204895, "learning_rate": 5.825215379385327e-05, "loss": 0.0219, "step": 37990 }, { "grad_norm": 0.316715806722641, "learning_rate": 5.823176815531057e-05, "loss": 0.044, "step": 38000 }, { "grad_norm": 0.27907586097717285, "learning_rate": 5.8211381110183826e-05, "loss": 0.0311, "step": 38010 }, { "grad_norm": 0.39413759112358093, "learning_rate": 5.8190992661956645e-05, "loss": 0.0226, "step": 38020 }, { "grad_norm": 0.31248533725738525, "learning_rate": 5.817060281411284e-05, "loss": 0.0224, "step": 38030 }, { "grad_norm": 0.2681208848953247, "learning_rate": 5.815021157013647e-05, "loss": 0.0241, "step": 38040 }, { "grad_norm": 0.3550228774547577, "learning_rate": 5.8129818933511856e-05, "loss": 0.0343, "step": 38050 }, { "grad_norm": 0.19261574745178223, "learning_rate": 5.8109424907723544e-05, "loss": 0.0276, "step": 38060 }, { "grad_norm": 0.2843894362449646, "learning_rate": 5.80890294962563e-05, "loss": 0.0242, "step": 38070 }, { "grad_norm": 0.3880428671836853, "learning_rate": 5.806863270259515e-05, "loss": 0.0295, "step": 38080 }, { "grad_norm": 0.3059903085231781, "learning_rate": 5.804823453022536e-05, "loss": 0.0228, "step": 38090 }, { "grad_norm": 0.36737772822380066, "learning_rate": 5.80278349826324e-05, "loss": 0.0284, "step": 38100 }, { "grad_norm": 0.3000308573246002, "learning_rate": 5.8007434063302014e-05, "loss": 0.0228, "step": 38110 }, { "grad_norm": 0.41299763321876526, "learning_rate": 5.7987031775720136e-05, "loss": 0.027, "step": 38120 }, { "grad_norm": 0.4191504418849945, "learning_rate": 5.7966628123372976e-05, "loss": 0.0311, "step": 38130 }, { "grad_norm": 0.5857074856758118, "learning_rate": 5.7946223109746956e-05, "loss": 0.0373, "step": 38140 }, { "grad_norm": 0.34373363852500916, "learning_rate": 5.7925816738328754e-05, "loss": 0.0217, "step": 38150 }, { "grad_norm": 0.35555437207221985, "learning_rate": 5.790540901260521e-05, "loss": 0.0284, "step": 38160 }, { "grad_norm": 0.36637094616889954, "learning_rate": 5.788499993606351e-05, "loss": 0.0249, "step": 38170 }, { "grad_norm": 0.2841246426105499, "learning_rate": 5.786458951219096e-05, "loss": 0.0274, "step": 38180 }, { "grad_norm": 0.3039941191673279, "learning_rate": 5.784417774447517e-05, "loss": 0.0338, "step": 38190 }, { "grad_norm": 0.23098665475845337, "learning_rate": 5.782376463640393e-05, "loss": 0.0266, "step": 38200 }, { "grad_norm": 0.366140753030777, "learning_rate": 5.780335019146531e-05, "loss": 0.0258, "step": 38210 }, { "grad_norm": 0.3024696111679077, "learning_rate": 5.778293441314755e-05, "loss": 0.0247, "step": 38220 }, { "grad_norm": 0.35446247458457947, "learning_rate": 5.776251730493917e-05, "loss": 0.0268, "step": 38230 }, { "grad_norm": 0.25677919387817383, "learning_rate": 5.774209887032887e-05, "loss": 0.0269, "step": 38240 }, { "grad_norm": 0.3305968642234802, "learning_rate": 5.772167911280565e-05, "loss": 0.0223, "step": 38250 }, { "grad_norm": 0.2821747958660126, "learning_rate": 5.770125803585864e-05, "loss": 0.0252, "step": 38260 }, { "grad_norm": 0.30620166659355164, "learning_rate": 5.768083564297726e-05, "loss": 0.0316, "step": 38270 }, { "grad_norm": 0.4032452404499054, "learning_rate": 5.766041193765114e-05, "loss": 0.0274, "step": 38280 }, { "grad_norm": 0.34983357787132263, "learning_rate": 5.763998692337015e-05, "loss": 0.021, "step": 38290 }, { "grad_norm": 0.36286503076553345, "learning_rate": 5.761956060362433e-05, "loss": 0.0255, "step": 38300 }, { "grad_norm": 0.3172549307346344, "learning_rate": 5.7599132981904e-05, "loss": 0.028, "step": 38310 }, { "grad_norm": 0.3397781550884247, "learning_rate": 5.75787040616997e-05, "loss": 0.0284, "step": 38320 }, { "grad_norm": 0.3444603979587555, "learning_rate": 5.755827384650212e-05, "loss": 0.0266, "step": 38330 }, { "grad_norm": 0.28596311807632446, "learning_rate": 5.753784233980228e-05, "loss": 0.0262, "step": 38340 }, { "grad_norm": 0.1864202469587326, "learning_rate": 5.751740954509135e-05, "loss": 0.0196, "step": 38350 }, { "grad_norm": 0.1649155616760254, "learning_rate": 5.7496975465860715e-05, "loss": 0.0156, "step": 38360 }, { "grad_norm": 0.33441415429115295, "learning_rate": 5.747654010560202e-05, "loss": 0.022, "step": 38370 }, { "grad_norm": 0.3009643256664276, "learning_rate": 5.7456103467807097e-05, "loss": 0.0279, "step": 38380 }, { "grad_norm": 0.3205961287021637, "learning_rate": 5.7435665555968046e-05, "loss": 0.0187, "step": 38390 }, { "grad_norm": 0.29154086112976074, "learning_rate": 5.74152263735771e-05, "loss": 0.0301, "step": 38400 }, { "grad_norm": 0.37435704469680786, "learning_rate": 5.739478592412677e-05, "loss": 0.0197, "step": 38410 }, { "grad_norm": 0.6096300482749939, "learning_rate": 5.7374344211109766e-05, "loss": 0.0349, "step": 38420 }, { "grad_norm": 0.4385960102081299, "learning_rate": 5.735390123801904e-05, "loss": 0.0268, "step": 38430 }, { "grad_norm": 0.46744880080223083, "learning_rate": 5.7333457008347704e-05, "loss": 0.0286, "step": 38440 }, { "grad_norm": 0.35121992230415344, "learning_rate": 5.7313011525589156e-05, "loss": 0.0283, "step": 38450 }, { "grad_norm": 0.24426232278347015, "learning_rate": 5.729256479323694e-05, "loss": 0.025, "step": 38460 }, { "grad_norm": 0.4732781946659088, "learning_rate": 5.727211681478485e-05, "loss": 0.0246, "step": 38470 }, { "grad_norm": 0.229282945394516, "learning_rate": 5.7251667593726886e-05, "loss": 0.0293, "step": 38480 }, { "grad_norm": 0.257766455411911, "learning_rate": 5.723121713355728e-05, "loss": 0.0324, "step": 38490 }, { "grad_norm": 0.2701753079891205, "learning_rate": 5.721076543777044e-05, "loss": 0.0276, "step": 38500 }, { "grad_norm": 0.3563632071018219, "learning_rate": 5.7190312509860986e-05, "loss": 0.021, "step": 38510 }, { "grad_norm": 0.3320244252681732, "learning_rate": 5.716985835332379e-05, "loss": 0.0231, "step": 38520 }, { "grad_norm": 0.30707064270973206, "learning_rate": 5.714940297165389e-05, "loss": 0.0245, "step": 38530 }, { "grad_norm": 0.2592586278915405, "learning_rate": 5.712894636834656e-05, "loss": 0.027, "step": 38540 }, { "grad_norm": 0.40250056982040405, "learning_rate": 5.7108488546897276e-05, "loss": 0.0207, "step": 38550 }, { "grad_norm": 0.19177387654781342, "learning_rate": 5.708802951080172e-05, "loss": 0.02, "step": 38560 }, { "grad_norm": 0.40249523520469666, "learning_rate": 5.706756926355576e-05, "loss": 0.0274, "step": 38570 }, { "grad_norm": 0.2727321684360504, "learning_rate": 5.704710780865554e-05, "loss": 0.0338, "step": 38580 }, { "grad_norm": 0.28606560826301575, "learning_rate": 5.7026645149597325e-05, "loss": 0.0263, "step": 38590 }, { "grad_norm": 0.4105091989040375, "learning_rate": 5.700618128987764e-05, "loss": 0.0357, "step": 38600 }, { "grad_norm": 0.20324382185935974, "learning_rate": 5.698571623299317e-05, "loss": 0.0185, "step": 38610 }, { "grad_norm": 0.26855650544166565, "learning_rate": 5.696524998244086e-05, "loss": 0.0201, "step": 38620 }, { "grad_norm": 0.24169263243675232, "learning_rate": 5.6944782541717836e-05, "loss": 0.0295, "step": 38630 }, { "grad_norm": 0.31010013818740845, "learning_rate": 5.69243139143214e-05, "loss": 0.019, "step": 38640 }, { "grad_norm": 0.28046804666519165, "learning_rate": 5.6903844103749125e-05, "loss": 0.0213, "step": 38650 }, { "grad_norm": 0.29517534375190735, "learning_rate": 5.688337311349869e-05, "loss": 0.0225, "step": 38660 }, { "grad_norm": 0.34821271896362305, "learning_rate": 5.6862900947068074e-05, "loss": 0.027, "step": 38670 }, { "grad_norm": 0.41567087173461914, "learning_rate": 5.6842427607955374e-05, "loss": 0.0274, "step": 38680 }, { "grad_norm": 0.23642376065254211, "learning_rate": 5.682195309965893e-05, "loss": 0.0203, "step": 38690 }, { "grad_norm": 0.45745331048965454, "learning_rate": 5.6801477425677294e-05, "loss": 0.0291, "step": 38700 }, { "grad_norm": 0.36299195885658264, "learning_rate": 5.678100058950917e-05, "loss": 0.0288, "step": 38710 }, { "grad_norm": 0.5518814325332642, "learning_rate": 5.676052259465352e-05, "loss": 0.0295, "step": 38720 }, { "grad_norm": 0.5668164491653442, "learning_rate": 5.674004344460945e-05, "loss": 0.0246, "step": 38730 }, { "grad_norm": 0.2032892256975174, "learning_rate": 5.6719563142876295e-05, "loss": 0.0201, "step": 38740 }, { "grad_norm": 0.2647833526134491, "learning_rate": 5.669908169295359e-05, "loss": 0.0225, "step": 38750 }, { "grad_norm": 0.42355915904045105, "learning_rate": 5.667859909834105e-05, "loss": 0.0264, "step": 38760 }, { "grad_norm": 0.29005441069602966, "learning_rate": 5.6658115362538565e-05, "loss": 0.0255, "step": 38770 }, { "grad_norm": 0.3549552261829376, "learning_rate": 5.663763048904628e-05, "loss": 0.0209, "step": 38780 }, { "grad_norm": 0.33835214376449585, "learning_rate": 5.661714448136447e-05, "loss": 0.0195, "step": 38790 }, { "grad_norm": 0.22369599342346191, "learning_rate": 5.659665734299366e-05, "loss": 0.0201, "step": 38800 }, { "grad_norm": 0.31663334369659424, "learning_rate": 5.6576169077434516e-05, "loss": 0.0249, "step": 38810 }, { "grad_norm": 0.26218676567077637, "learning_rate": 5.6555679688187944e-05, "loss": 0.0275, "step": 38820 }, { "grad_norm": 0.5015304088592529, "learning_rate": 5.6535189178755e-05, "loss": 0.0296, "step": 38830 }, { "grad_norm": 0.29097115993499756, "learning_rate": 5.651469755263698e-05, "loss": 0.0308, "step": 38840 }, { "grad_norm": 0.16807213425636292, "learning_rate": 5.6494204813335316e-05, "loss": 0.0316, "step": 38850 }, { "grad_norm": 0.2784346342086792, "learning_rate": 5.647371096435168e-05, "loss": 0.0193, "step": 38860 }, { "grad_norm": 0.37605535984039307, "learning_rate": 5.645321600918788e-05, "loss": 0.0157, "step": 38870 }, { "grad_norm": 0.4052322208881378, "learning_rate": 5.643271995134597e-05, "loss": 0.0313, "step": 38880 }, { "grad_norm": 0.2392929345369339, "learning_rate": 5.641222279432814e-05, "loss": 0.0243, "step": 38890 }, { "grad_norm": 0.3517214357852936, "learning_rate": 5.6391724541636834e-05, "loss": 0.0217, "step": 38900 }, { "grad_norm": 0.3461625576019287, "learning_rate": 5.6371225196774605e-05, "loss": 0.0296, "step": 38910 }, { "grad_norm": 0.3942202925682068, "learning_rate": 5.635072476324423e-05, "loss": 0.0281, "step": 38920 }, { "grad_norm": 0.22102299332618713, "learning_rate": 5.63302232445487e-05, "loss": 0.0178, "step": 38930 }, { "grad_norm": 0.3449041247367859, "learning_rate": 5.6309720644191144e-05, "loss": 0.0301, "step": 38940 }, { "grad_norm": 0.46516674757003784, "learning_rate": 5.628921696567491e-05, "loss": 0.0272, "step": 38950 }, { "grad_norm": 0.2318221777677536, "learning_rate": 5.62687122125035e-05, "loss": 0.0233, "step": 38960 }, { "grad_norm": 0.3272753655910492, "learning_rate": 5.624820638818062e-05, "loss": 0.0242, "step": 38970 }, { "grad_norm": 0.4197727143764496, "learning_rate": 5.6227699496210164e-05, "loss": 0.0372, "step": 38980 }, { "grad_norm": 0.3363933861255646, "learning_rate": 5.6207191540096195e-05, "loss": 0.0269, "step": 38990 }, { "grad_norm": 0.3966897428035736, "learning_rate": 5.618668252334296e-05, "loss": 0.0271, "step": 39000 }, { "grad_norm": 0.4017120897769928, "learning_rate": 5.616617244945488e-05, "loss": 0.0227, "step": 39010 }, { "grad_norm": 0.29105591773986816, "learning_rate": 5.614566132193656e-05, "loss": 0.0224, "step": 39020 }, { "grad_norm": 0.2746066749095917, "learning_rate": 5.612514914429282e-05, "loss": 0.032, "step": 39030 }, { "grad_norm": 0.3280901610851288, "learning_rate": 5.610463592002863e-05, "loss": 0.0233, "step": 39040 }, { "grad_norm": 0.3912464380264282, "learning_rate": 5.608412165264909e-05, "loss": 0.0256, "step": 39050 }, { "grad_norm": 0.20799191296100616, "learning_rate": 5.606360634565959e-05, "loss": 0.0175, "step": 39060 }, { "grad_norm": 0.2526998817920685, "learning_rate": 5.604309000256559e-05, "loss": 0.0224, "step": 39070 }, { "grad_norm": 0.33189889788627625, "learning_rate": 5.6022572626872785e-05, "loss": 0.0215, "step": 39080 }, { "grad_norm": 0.25196516513824463, "learning_rate": 5.600205422208704e-05, "loss": 0.0236, "step": 39090 }, { "grad_norm": 0.34629836678504944, "learning_rate": 5.59815347917144e-05, "loss": 0.0285, "step": 39100 }, { "grad_norm": 0.37296023964881897, "learning_rate": 5.596101433926103e-05, "loss": 0.0211, "step": 39110 }, { "grad_norm": 0.3111063539981842, "learning_rate": 5.5940492868233364e-05, "loss": 0.0167, "step": 39120 }, { "grad_norm": 0.2668570578098297, "learning_rate": 5.591997038213793e-05, "loss": 0.0229, "step": 39130 }, { "grad_norm": 0.39579784870147705, "learning_rate": 5.5899446884481475e-05, "loss": 0.0221, "step": 39140 }, { "grad_norm": 0.3218860626220703, "learning_rate": 5.5878922378770906e-05, "loss": 0.0361, "step": 39150 }, { "grad_norm": 0.2529977262020111, "learning_rate": 5.5858396868513285e-05, "loss": 0.0257, "step": 39160 }, { "grad_norm": 0.33004656434059143, "learning_rate": 5.583787035721586e-05, "loss": 0.0208, "step": 39170 }, { "grad_norm": 0.2682110071182251, "learning_rate": 5.581734284838606e-05, "loss": 0.0212, "step": 39180 }, { "grad_norm": 0.4081844091415405, "learning_rate": 5.579681434553147e-05, "loss": 0.024, "step": 39190 }, { "grad_norm": 0.20470108091831207, "learning_rate": 5.5776284852159854e-05, "loss": 0.0169, "step": 39200 }, { "grad_norm": 0.2750053107738495, "learning_rate": 5.575575437177913e-05, "loss": 0.0273, "step": 39210 }, { "grad_norm": 0.36526864767074585, "learning_rate": 5.573522290789742e-05, "loss": 0.0194, "step": 39220 }, { "grad_norm": 0.24754804372787476, "learning_rate": 5.571469046402298e-05, "loss": 0.0299, "step": 39230 }, { "grad_norm": 0.310217946767807, "learning_rate": 5.5694157043664205e-05, "loss": 0.0256, "step": 39240 }, { "grad_norm": 0.17047348618507385, "learning_rate": 5.567362265032975e-05, "loss": 0.0153, "step": 39250 }, { "grad_norm": 0.3115096390247345, "learning_rate": 5.565308728752836e-05, "loss": 0.0225, "step": 39260 }, { "grad_norm": 0.29074928164482117, "learning_rate": 5.5632550958768956e-05, "loss": 0.0216, "step": 39270 }, { "grad_norm": 0.42583146691322327, "learning_rate": 5.5612013667560636e-05, "loss": 0.0234, "step": 39280 }, { "grad_norm": 0.37660276889801025, "learning_rate": 5.5591475417412676e-05, "loss": 0.0356, "step": 39290 }, { "grad_norm": 0.3886454701423645, "learning_rate": 5.557093621183451e-05, "loss": 0.028, "step": 39300 }, { "grad_norm": 0.28094422817230225, "learning_rate": 5.55503960543357e-05, "loss": 0.0245, "step": 39310 }, { "grad_norm": 0.371180921792984, "learning_rate": 5.552985494842601e-05, "loss": 0.0297, "step": 39320 }, { "grad_norm": 0.2712882459163666, "learning_rate": 5.550931289761534e-05, "loss": 0.0229, "step": 39330 }, { "grad_norm": 0.35827240347862244, "learning_rate": 5.548876990541378e-05, "loss": 0.0222, "step": 39340 }, { "grad_norm": 0.4067690074443817, "learning_rate": 5.5468225975331564e-05, "loss": 0.0288, "step": 39350 }, { "grad_norm": 0.3230920135974884, "learning_rate": 5.544768111087909e-05, "loss": 0.0234, "step": 39360 }, { "grad_norm": 0.25933727622032166, "learning_rate": 5.54271353155669e-05, "loss": 0.0297, "step": 39370 }, { "grad_norm": 0.34148165583610535, "learning_rate": 5.5406588592905715e-05, "loss": 0.0263, "step": 39380 }, { "grad_norm": 0.24144192039966583, "learning_rate": 5.5386040946406416e-05, "loss": 0.0305, "step": 39390 }, { "grad_norm": 0.45770397782325745, "learning_rate": 5.536549237958004e-05, "loss": 0.0288, "step": 39400 }, { "grad_norm": 0.23381690680980682, "learning_rate": 5.5344942895937744e-05, "loss": 0.0361, "step": 39410 }, { "grad_norm": 0.2719796597957611, "learning_rate": 5.5324392498990904e-05, "loss": 0.0187, "step": 39420 }, { "grad_norm": 0.4361991882324219, "learning_rate": 5.5303841192251e-05, "loss": 0.0216, "step": 39430 }, { "grad_norm": 0.3730636239051819, "learning_rate": 5.52832889792297e-05, "loss": 0.0275, "step": 39440 }, { "grad_norm": 0.30011501908302307, "learning_rate": 5.526273586343881e-05, "loss": 0.0239, "step": 39450 }, { "grad_norm": 0.3416365087032318, "learning_rate": 5.5242181848390306e-05, "loss": 0.0213, "step": 39460 }, { "grad_norm": 0.2939034104347229, "learning_rate": 5.5221626937596285e-05, "loss": 0.0167, "step": 39470 }, { "grad_norm": 0.40960001945495605, "learning_rate": 5.520107113456903e-05, "loss": 0.0337, "step": 39480 }, { "grad_norm": 0.45521801710128784, "learning_rate": 5.5180514442820974e-05, "loss": 0.0376, "step": 39490 }, { "grad_norm": 0.4801931083202362, "learning_rate": 5.515995686586469e-05, "loss": 0.0217, "step": 39500 }, { "grad_norm": 0.37807705998420715, "learning_rate": 5.5139398407212916e-05, "loss": 0.0224, "step": 39510 }, { "grad_norm": 0.3063068985939026, "learning_rate": 5.511883907037849e-05, "loss": 0.0247, "step": 39520 }, { "grad_norm": 0.4985203444957733, "learning_rate": 5.509827885887449e-05, "loss": 0.0273, "step": 39530 }, { "grad_norm": 0.37467366456985474, "learning_rate": 5.507771777621406e-05, "loss": 0.0236, "step": 39540 }, { "grad_norm": 0.19433407485485077, "learning_rate": 5.505715582591052e-05, "loss": 0.0307, "step": 39550 }, { "grad_norm": 0.3290586769580841, "learning_rate": 5.50365930114774e-05, "loss": 0.0287, "step": 39560 }, { "grad_norm": 0.2796137034893036, "learning_rate": 5.5016029336428255e-05, "loss": 0.021, "step": 39570 }, { "grad_norm": 0.3439214825630188, "learning_rate": 5.49954648042769e-05, "loss": 0.0271, "step": 39580 }, { "grad_norm": 0.37920504808425903, "learning_rate": 5.4974899418537226e-05, "loss": 0.0341, "step": 39590 }, { "grad_norm": 0.6811408400535583, "learning_rate": 5.4954333182723316e-05, "loss": 0.0244, "step": 39600 }, { "grad_norm": 0.44135841727256775, "learning_rate": 5.493376610034937e-05, "loss": 0.0245, "step": 39610 }, { "grad_norm": 0.3084084093570709, "learning_rate": 5.4913198174929735e-05, "loss": 0.0218, "step": 39620 }, { "grad_norm": 0.22472545504570007, "learning_rate": 5.489262940997889e-05, "loss": 0.029, "step": 39630 }, { "grad_norm": 0.18735148012638092, "learning_rate": 5.487205980901151e-05, "loss": 0.0256, "step": 39640 }, { "grad_norm": 0.4571046233177185, "learning_rate": 5.485148937554234e-05, "loss": 0.0433, "step": 39650 }, { "grad_norm": 0.2582273483276367, "learning_rate": 5.483091811308635e-05, "loss": 0.0253, "step": 39660 }, { "grad_norm": 0.3357565402984619, "learning_rate": 5.4810346025158564e-05, "loss": 0.0245, "step": 39670 }, { "grad_norm": 0.2756047248840332, "learning_rate": 5.478977311527421e-05, "loss": 0.0257, "step": 39680 }, { "grad_norm": 0.20650745928287506, "learning_rate": 5.476919938694863e-05, "loss": 0.0241, "step": 39690 }, { "grad_norm": 0.2653321921825409, "learning_rate": 5.474862484369733e-05, "loss": 0.0187, "step": 39700 }, { "grad_norm": 0.22034260630607605, "learning_rate": 5.472804948903589e-05, "loss": 0.0232, "step": 39710 }, { "grad_norm": 0.3157701790332794, "learning_rate": 5.470747332648013e-05, "loss": 0.0194, "step": 39720 }, { "grad_norm": 0.3642890453338623, "learning_rate": 5.468689635954591e-05, "loss": 0.0367, "step": 39730 }, { "grad_norm": 0.40941959619522095, "learning_rate": 5.46663185917493e-05, "loss": 0.0267, "step": 39740 }, { "grad_norm": 0.33394739031791687, "learning_rate": 5.464574002660645e-05, "loss": 0.0242, "step": 39750 }, { "grad_norm": 0.18968020379543304, "learning_rate": 5.4625160667633724e-05, "loss": 0.0286, "step": 39760 }, { "grad_norm": 0.180820032954216, "learning_rate": 5.4604580518347505e-05, "loss": 0.0287, "step": 39770 }, { "grad_norm": 0.32339078187942505, "learning_rate": 5.458399958226442e-05, "loss": 0.0246, "step": 39780 }, { "grad_norm": 0.230246901512146, "learning_rate": 5.456341786290119e-05, "loss": 0.0242, "step": 39790 }, { "grad_norm": 0.3630484342575073, "learning_rate": 5.454283536377465e-05, "loss": 0.0313, "step": 39800 }, { "grad_norm": 0.37235134840011597, "learning_rate": 5.452225208840179e-05, "loss": 0.0317, "step": 39810 }, { "grad_norm": 0.32209768891334534, "learning_rate": 5.450166804029975e-05, "loss": 0.0357, "step": 39820 }, { "grad_norm": 0.7197932004928589, "learning_rate": 5.448108322298574e-05, "loss": 0.022, "step": 39830 }, { "grad_norm": 0.19524164497852325, "learning_rate": 5.446049763997717e-05, "loss": 0.0205, "step": 39840 }, { "grad_norm": 0.4565868377685547, "learning_rate": 5.4439911294791546e-05, "loss": 0.0294, "step": 39850 }, { "grad_norm": 0.35006335377693176, "learning_rate": 5.441932419094652e-05, "loss": 0.0269, "step": 39860 }, { "grad_norm": 0.1703166514635086, "learning_rate": 5.439873633195985e-05, "loss": 0.0196, "step": 39870 }, { "grad_norm": 0.24533575773239136, "learning_rate": 5.437814772134947e-05, "loss": 0.0289, "step": 39880 }, { "grad_norm": 0.33045801520347595, "learning_rate": 5.4357558362633366e-05, "loss": 0.0235, "step": 39890 }, { "grad_norm": 0.20655648410320282, "learning_rate": 5.4336968259329726e-05, "loss": 0.0181, "step": 39900 }, { "grad_norm": 0.2894490659236908, "learning_rate": 5.431637741495681e-05, "loss": 0.0274, "step": 39910 }, { "grad_norm": 0.5513847470283508, "learning_rate": 5.429578583303307e-05, "loss": 0.0277, "step": 39920 }, { "grad_norm": 0.3010338544845581, "learning_rate": 5.427519351707701e-05, "loss": 0.0242, "step": 39930 }, { "grad_norm": 0.17758293449878693, "learning_rate": 5.42546004706073e-05, "loss": 0.0191, "step": 39940 }, { "grad_norm": 0.32904520630836487, "learning_rate": 5.4234006697142735e-05, "loss": 0.0226, "step": 39950 }, { "grad_norm": 0.21690961718559265, "learning_rate": 5.421341220020224e-05, "loss": 0.0223, "step": 39960 }, { "grad_norm": 0.30004367232322693, "learning_rate": 5.419281698330482e-05, "loss": 0.0249, "step": 39970 }, { "grad_norm": 0.19512778520584106, "learning_rate": 5.4172221049969665e-05, "loss": 0.0216, "step": 39980 }, { "grad_norm": 0.3188093304634094, "learning_rate": 5.415162440371604e-05, "loss": 0.0209, "step": 39990 }, { "grad_norm": 0.3192623257637024, "learning_rate": 5.413102704806334e-05, "loss": 0.0267, "step": 40000 }, { "grad_norm": 0.24307937920093536, "learning_rate": 5.41104289865311e-05, "loss": 0.0251, "step": 40010 }, { "grad_norm": 0.2969112694263458, "learning_rate": 5.408983022263898e-05, "loss": 0.0197, "step": 40020 }, { "grad_norm": 0.36614859104156494, "learning_rate": 5.406923075990673e-05, "loss": 0.029, "step": 40030 }, { "grad_norm": 0.1988174170255661, "learning_rate": 5.404863060185423e-05, "loss": 0.02, "step": 40040 }, { "grad_norm": 0.4447190463542938, "learning_rate": 5.40280297520015e-05, "loss": 0.0285, "step": 40050 }, { "grad_norm": 0.3268045485019684, "learning_rate": 5.400742821386865e-05, "loss": 0.0198, "step": 40060 }, { "grad_norm": 0.2635117471218109, "learning_rate": 5.398682599097592e-05, "loss": 0.0255, "step": 40070 }, { "grad_norm": 0.2707625925540924, "learning_rate": 5.396622308684367e-05, "loss": 0.0242, "step": 40080 }, { "grad_norm": 0.42980167269706726, "learning_rate": 5.394561950499236e-05, "loss": 0.0312, "step": 40090 }, { "grad_norm": 0.26014310121536255, "learning_rate": 5.39250152489426e-05, "loss": 0.028, "step": 40100 }, { "grad_norm": 0.2708563208580017, "learning_rate": 5.390441032221507e-05, "loss": 0.0199, "step": 40110 }, { "grad_norm": 0.3306858241558075, "learning_rate": 5.388380472833062e-05, "loss": 0.0266, "step": 40120 }, { "grad_norm": 0.19768072664737701, "learning_rate": 5.386319847081016e-05, "loss": 0.0197, "step": 40130 }, { "grad_norm": 0.4414941966533661, "learning_rate": 5.384259155317473e-05, "loss": 0.0292, "step": 40140 }, { "grad_norm": 0.3663599491119385, "learning_rate": 5.38219839789455e-05, "loss": 0.028, "step": 40150 }, { "grad_norm": 0.34523841738700867, "learning_rate": 5.380137575164376e-05, "loss": 0.0348, "step": 40160 }, { "grad_norm": 0.22511395812034607, "learning_rate": 5.378076687479085e-05, "loss": 0.0227, "step": 40170 }, { "grad_norm": 0.30213016271591187, "learning_rate": 5.3760157351908305e-05, "loss": 0.0205, "step": 40180 }, { "grad_norm": 0.3581112027168274, "learning_rate": 5.373954718651768e-05, "loss": 0.0305, "step": 40190 }, { "grad_norm": 0.21061331033706665, "learning_rate": 5.371893638214074e-05, "loss": 0.0199, "step": 40200 }, { "grad_norm": 0.303087055683136, "learning_rate": 5.369832494229927e-05, "loss": 0.0206, "step": 40210 }, { "grad_norm": 0.3732828199863434, "learning_rate": 5.367771287051524e-05, "loss": 0.0259, "step": 40220 }, { "grad_norm": 0.16304820775985718, "learning_rate": 5.365710017031065e-05, "loss": 0.0191, "step": 40230 }, { "grad_norm": 0.3702610731124878, "learning_rate": 5.3636486845207654e-05, "loss": 0.0345, "step": 40240 }, { "grad_norm": 0.3291376829147339, "learning_rate": 5.3615872898728526e-05, "loss": 0.0205, "step": 40250 }, { "grad_norm": 0.39554399251937866, "learning_rate": 5.3595258334395614e-05, "loss": 0.0252, "step": 40260 }, { "grad_norm": 0.3081952929496765, "learning_rate": 5.3574643155731396e-05, "loss": 0.0307, "step": 40270 }, { "grad_norm": 0.3110429346561432, "learning_rate": 5.3554027366258405e-05, "loss": 0.0203, "step": 40280 }, { "grad_norm": 0.3349902629852295, "learning_rate": 5.3533410969499355e-05, "loss": 0.0256, "step": 40290 }, { "grad_norm": 0.18108369410037994, "learning_rate": 5.3512793968977e-05, "loss": 0.022, "step": 40300 }, { "grad_norm": 0.42514315247535706, "learning_rate": 5.349217636821423e-05, "loss": 0.0262, "step": 40310 }, { "grad_norm": 0.37098318338394165, "learning_rate": 5.347155817073405e-05, "loss": 0.0219, "step": 40320 }, { "grad_norm": 0.2491072416305542, "learning_rate": 5.34509393800595e-05, "loss": 0.0215, "step": 40330 }, { "grad_norm": 0.2705603539943695, "learning_rate": 5.343031999971381e-05, "loss": 0.0217, "step": 40340 }, { "grad_norm": 0.2683572471141815, "learning_rate": 5.340970003322026e-05, "loss": 0.0276, "step": 40350 }, { "grad_norm": 0.35982373356819153, "learning_rate": 5.33890794841022e-05, "loss": 0.0204, "step": 40360 }, { "grad_norm": 0.23864220082759857, "learning_rate": 5.336845835588318e-05, "loss": 0.0232, "step": 40370 }, { "grad_norm": 0.36032426357269287, "learning_rate": 5.334783665208674e-05, "loss": 0.0281, "step": 40380 }, { "grad_norm": 0.32731592655181885, "learning_rate": 5.332721437623657e-05, "loss": 0.0232, "step": 40390 }, { "grad_norm": 0.24393923580646515, "learning_rate": 5.3306591531856464e-05, "loss": 0.0247, "step": 40400 }, { "grad_norm": 0.24388043582439423, "learning_rate": 5.3285968122470295e-05, "loss": 0.0179, "step": 40410 }, { "grad_norm": 0.2051643431186676, "learning_rate": 5.326534415160207e-05, "loss": 0.0266, "step": 40420 }, { "grad_norm": 0.21837450563907623, "learning_rate": 5.324471962277582e-05, "loss": 0.0291, "step": 40430 }, { "grad_norm": 0.2813267111778259, "learning_rate": 5.3224094539515746e-05, "loss": 0.0258, "step": 40440 }, { "grad_norm": 0.22724217176437378, "learning_rate": 5.3203468905346076e-05, "loss": 0.0209, "step": 40450 }, { "grad_norm": 0.35688942670822144, "learning_rate": 5.3182842723791195e-05, "loss": 0.0262, "step": 40460 }, { "grad_norm": 0.40858328342437744, "learning_rate": 5.316221599837554e-05, "loss": 0.0274, "step": 40470 }, { "grad_norm": 0.3507164418697357, "learning_rate": 5.314158873262366e-05, "loss": 0.0334, "step": 40480 }, { "grad_norm": 0.28224676847457886, "learning_rate": 5.312096093006018e-05, "loss": 0.0201, "step": 40490 }, { "grad_norm": 0.3376398980617523, "learning_rate": 5.3100332594209846e-05, "loss": 0.025, "step": 40500 }, { "grad_norm": 0.2860953211784363, "learning_rate": 5.3079703728597454e-05, "loss": 0.0262, "step": 40510 }, { "grad_norm": 0.30509209632873535, "learning_rate": 5.305907433674794e-05, "loss": 0.0232, "step": 40520 }, { "grad_norm": 0.25284144282341003, "learning_rate": 5.3038444422186284e-05, "loss": 0.0217, "step": 40530 }, { "grad_norm": 0.3463686406612396, "learning_rate": 5.301781398843757e-05, "loss": 0.0179, "step": 40540 }, { "grad_norm": 0.37033864855766296, "learning_rate": 5.299718303902699e-05, "loss": 0.0277, "step": 40550 }, { "grad_norm": 0.18497565388679504, "learning_rate": 5.297655157747979e-05, "loss": 0.0272, "step": 40560 }, { "grad_norm": 0.3395443260669708, "learning_rate": 5.295591960732136e-05, "loss": 0.0218, "step": 40570 }, { "grad_norm": 0.23881196975708008, "learning_rate": 5.293528713207708e-05, "loss": 0.0241, "step": 40580 }, { "grad_norm": 1.2915046215057373, "learning_rate": 5.291465415527253e-05, "loss": 0.0377, "step": 40590 }, { "grad_norm": 0.4516933858394623, "learning_rate": 5.2894020680433296e-05, "loss": 0.0241, "step": 40600 }, { "grad_norm": 0.33928540349006653, "learning_rate": 5.287338671108507e-05, "loss": 0.0226, "step": 40610 }, { "grad_norm": 0.4426233768463135, "learning_rate": 5.285275225075367e-05, "loss": 0.0279, "step": 40620 }, { "grad_norm": 0.3013686537742615, "learning_rate": 5.283211730296492e-05, "loss": 0.0238, "step": 40630 }, { "grad_norm": 0.2562166452407837, "learning_rate": 5.281148187124477e-05, "loss": 0.0244, "step": 40640 }, { "grad_norm": 0.4561983346939087, "learning_rate": 5.279084595911927e-05, "loss": 0.0206, "step": 40650 }, { "grad_norm": 0.4228198230266571, "learning_rate": 5.277020957011453e-05, "loss": 0.0359, "step": 40660 }, { "grad_norm": 0.40259793400764465, "learning_rate": 5.274957270775673e-05, "loss": 0.0295, "step": 40670 }, { "grad_norm": 0.23432143032550812, "learning_rate": 5.2728935375572164e-05, "loss": 0.019, "step": 40680 }, { "grad_norm": 0.26221346855163574, "learning_rate": 5.2708297577087176e-05, "loss": 0.0248, "step": 40690 }, { "grad_norm": 0.5010645389556885, "learning_rate": 5.2687659315828196e-05, "loss": 0.0226, "step": 40700 }, { "grad_norm": 0.24727478623390198, "learning_rate": 5.266702059532175e-05, "loss": 0.0236, "step": 40710 }, { "grad_norm": 0.3406299352645874, "learning_rate": 5.264638141909444e-05, "loss": 0.0291, "step": 40720 }, { "grad_norm": 0.1903456747531891, "learning_rate": 5.2625741790672925e-05, "loss": 0.025, "step": 40730 }, { "grad_norm": 0.22501644492149353, "learning_rate": 5.2605101713583935e-05, "loss": 0.0215, "step": 40740 }, { "grad_norm": 0.4101041257381439, "learning_rate": 5.2584461191354315e-05, "loss": 0.0237, "step": 40750 }, { "grad_norm": 0.36405810713768005, "learning_rate": 5.2563820227510964e-05, "loss": 0.0178, "step": 40760 }, { "grad_norm": 0.25895488262176514, "learning_rate": 5.2543178825580864e-05, "loss": 0.0269, "step": 40770 }, { "grad_norm": 0.4310440421104431, "learning_rate": 5.2522536989091054e-05, "loss": 0.0294, "step": 40780 }, { "grad_norm": 0.3285765051841736, "learning_rate": 5.2501894721568655e-05, "loss": 0.0204, "step": 40790 }, { "grad_norm": 0.23578865826129913, "learning_rate": 5.248125202654089e-05, "loss": 0.0189, "step": 40800 }, { "grad_norm": 0.3493635058403015, "learning_rate": 5.246060890753501e-05, "loss": 0.0353, "step": 40810 }, { "grad_norm": 0.24884112179279327, "learning_rate": 5.243996536807837e-05, "loss": 0.0199, "step": 40820 }, { "grad_norm": 0.2822776436805725, "learning_rate": 5.2419321411698384e-05, "loss": 0.0223, "step": 40830 }, { "grad_norm": 0.33100104331970215, "learning_rate": 5.239867704192253e-05, "loss": 0.0237, "step": 40840 }, { "grad_norm": 0.13483953475952148, "learning_rate": 5.237803226227838e-05, "loss": 0.0181, "step": 40850 }, { "grad_norm": 0.2979872226715088, "learning_rate": 5.235738707629354e-05, "loss": 0.021, "step": 40860 }, { "grad_norm": 0.2318919599056244, "learning_rate": 5.233674148749575e-05, "loss": 0.0247, "step": 40870 }, { "grad_norm": 0.3524339497089386, "learning_rate": 5.231609549941272e-05, "loss": 0.0314, "step": 40880 }, { "grad_norm": 0.30692264437675476, "learning_rate": 5.229544911557233e-05, "loss": 0.0279, "step": 40890 }, { "grad_norm": 0.3262973725795746, "learning_rate": 5.227480233950246e-05, "loss": 0.0237, "step": 40900 }, { "grad_norm": 0.34777531027793884, "learning_rate": 5.22541551747311e-05, "loss": 0.023, "step": 40910 }, { "grad_norm": 0.3183707296848297, "learning_rate": 5.223350762478626e-05, "loss": 0.0183, "step": 40920 }, { "grad_norm": 0.17171794176101685, "learning_rate": 5.221285969319606e-05, "loss": 0.0206, "step": 40930 }, { "grad_norm": 0.21244335174560547, "learning_rate": 5.219221138348865e-05, "loss": 0.0196, "step": 40940 }, { "grad_norm": 0.2518891990184784, "learning_rate": 5.217156269919228e-05, "loss": 0.0219, "step": 40950 }, { "grad_norm": 0.2137375921010971, "learning_rate": 5.215091364383523e-05, "loss": 0.0225, "step": 40960 }, { "grad_norm": 0.45201441645622253, "learning_rate": 5.213026422094588e-05, "loss": 0.0295, "step": 40970 }, { "grad_norm": 0.3084559440612793, "learning_rate": 5.210961443405262e-05, "loss": 0.0276, "step": 40980 }, { "grad_norm": 0.2415335774421692, "learning_rate": 5.208896428668396e-05, "loss": 0.0194, "step": 40990 }, { "grad_norm": 0.43071654438972473, "learning_rate": 5.206831378236845e-05, "loss": 0.0309, "step": 41000 }, { "grad_norm": 0.26837632060050964, "learning_rate": 5.2047662924634666e-05, "loss": 0.0219, "step": 41010 }, { "grad_norm": 0.36467596888542175, "learning_rate": 5.20270117170113e-05, "loss": 0.0238, "step": 41020 }, { "grad_norm": 0.36752769351005554, "learning_rate": 5.200636016302707e-05, "loss": 0.0328, "step": 41030 }, { "grad_norm": 0.33604276180267334, "learning_rate": 5.1985708266210754e-05, "loss": 0.0285, "step": 41040 }, { "grad_norm": 0.37702110409736633, "learning_rate": 5.1965056030091206e-05, "loss": 0.0293, "step": 41050 }, { "grad_norm": 0.24917195737361908, "learning_rate": 5.194440345819732e-05, "loss": 0.0218, "step": 41060 }, { "grad_norm": 0.39333823323249817, "learning_rate": 5.1923750554058084e-05, "loss": 0.0247, "step": 41070 }, { "grad_norm": 0.35219842195510864, "learning_rate": 5.1903097321202476e-05, "loss": 0.0165, "step": 41080 }, { "grad_norm": 0.4039713144302368, "learning_rate": 5.18824437631596e-05, "loss": 0.0383, "step": 41090 }, { "grad_norm": 0.33862149715423584, "learning_rate": 5.186178988345856e-05, "loss": 0.0229, "step": 41100 }, { "grad_norm": 0.28860965371131897, "learning_rate": 5.184113568562855e-05, "loss": 0.0294, "step": 41110 }, { "grad_norm": 0.32188880443573, "learning_rate": 5.18204811731988e-05, "loss": 0.031, "step": 41120 }, { "grad_norm": 0.34607526659965515, "learning_rate": 5.179982634969862e-05, "loss": 0.0236, "step": 41130 }, { "grad_norm": 0.4239808917045593, "learning_rate": 5.177917121865734e-05, "loss": 0.0237, "step": 41140 }, { "grad_norm": 0.4042590260505676, "learning_rate": 5.1758515783604346e-05, "loss": 0.0208, "step": 41150 }, { "grad_norm": 0.3997862637042999, "learning_rate": 5.173786004806911e-05, "loss": 0.0305, "step": 41160 }, { "grad_norm": 0.4362463653087616, "learning_rate": 5.1717204015581135e-05, "loss": 0.029, "step": 41170 }, { "grad_norm": 0.29263442754745483, "learning_rate": 5.169654768966994e-05, "loss": 0.0222, "step": 41180 }, { "grad_norm": 0.31575554609298706, "learning_rate": 5.1675891073865156e-05, "loss": 0.0206, "step": 41190 }, { "grad_norm": 0.2626156210899353, "learning_rate": 5.1655234171696424e-05, "loss": 0.0186, "step": 41200 }, { "grad_norm": 0.3266502022743225, "learning_rate": 5.163457698669343e-05, "loss": 0.0169, "step": 41210 }, { "grad_norm": 0.30242007970809937, "learning_rate": 5.1613919522385936e-05, "loss": 0.02, "step": 41220 }, { "grad_norm": 0.3042711913585663, "learning_rate": 5.1593261782303746e-05, "loss": 0.0307, "step": 41230 }, { "grad_norm": 0.22989675402641296, "learning_rate": 5.157260376997669e-05, "loss": 0.0188, "step": 41240 }, { "grad_norm": 0.44263169169425964, "learning_rate": 5.155194548893466e-05, "loss": 0.0235, "step": 41250 }, { "grad_norm": 0.23687656223773956, "learning_rate": 5.153128694270759e-05, "loss": 0.0321, "step": 41260 }, { "grad_norm": 0.2575901746749878, "learning_rate": 5.151062813482548e-05, "loss": 0.0284, "step": 41270 }, { "grad_norm": 0.4426594078540802, "learning_rate": 5.148996906881832e-05, "loss": 0.0215, "step": 41280 }, { "grad_norm": 0.3192763328552246, "learning_rate": 5.1469309748216196e-05, "loss": 0.0232, "step": 41290 }, { "grad_norm": 0.41798895597457886, "learning_rate": 5.144865017654923e-05, "loss": 0.0237, "step": 41300 }, { "grad_norm": 0.3598385453224182, "learning_rate": 5.1427990357347556e-05, "loss": 0.0278, "step": 41310 }, { "grad_norm": 0.28834038972854614, "learning_rate": 5.140733029414139e-05, "loss": 0.025, "step": 41320 }, { "grad_norm": 0.25626248121261597, "learning_rate": 5.138666999046098e-05, "loss": 0.0238, "step": 41330 }, { "grad_norm": 0.288703054189682, "learning_rate": 5.136600944983658e-05, "loss": 0.019, "step": 41340 }, { "grad_norm": 0.2117709219455719, "learning_rate": 5.134534867579853e-05, "loss": 0.029, "step": 41350 }, { "grad_norm": 0.24945509433746338, "learning_rate": 5.1324687671877184e-05, "loss": 0.0201, "step": 41360 }, { "grad_norm": 0.45783358812332153, "learning_rate": 5.130402644160296e-05, "loss": 0.0293, "step": 41370 }, { "grad_norm": 0.2805935740470886, "learning_rate": 5.128336498850628e-05, "loss": 0.0222, "step": 41380 }, { "grad_norm": 0.16398344933986664, "learning_rate": 5.126270331611761e-05, "loss": 0.0222, "step": 41390 }, { "grad_norm": 0.6636231541633606, "learning_rate": 5.124204142796748e-05, "loss": 0.0282, "step": 41400 }, { "grad_norm": 0.26907554268836975, "learning_rate": 5.122137932758644e-05, "loss": 0.023, "step": 41410 }, { "grad_norm": 0.26273980736732483, "learning_rate": 5.1200717018505074e-05, "loss": 0.0251, "step": 41420 }, { "grad_norm": 0.2637920677661896, "learning_rate": 5.118005450425403e-05, "loss": 0.0198, "step": 41430 }, { "grad_norm": 0.227527916431427, "learning_rate": 5.115939178836392e-05, "loss": 0.016, "step": 41440 }, { "grad_norm": 0.3323678970336914, "learning_rate": 5.113872887436547e-05, "loss": 0.0308, "step": 41450 }, { "grad_norm": 0.35268497467041016, "learning_rate": 5.11180657657894e-05, "loss": 0.0256, "step": 41460 }, { "grad_norm": 0.24903295934200287, "learning_rate": 5.109740246616649e-05, "loss": 0.023, "step": 41470 }, { "grad_norm": 0.4744470715522766, "learning_rate": 5.10767389790275e-05, "loss": 0.0336, "step": 41480 }, { "grad_norm": 0.33431726694107056, "learning_rate": 5.1056075307903265e-05, "loss": 0.0261, "step": 41490 }, { "grad_norm": 0.3028043210506439, "learning_rate": 5.1035411456324634e-05, "loss": 0.0212, "step": 41500 }, { "grad_norm": 0.21724292635917664, "learning_rate": 5.1014747427822504e-05, "loss": 0.0273, "step": 41510 }, { "grad_norm": 0.36975252628326416, "learning_rate": 5.0994083225927804e-05, "loss": 0.0229, "step": 41520 }, { "grad_norm": 0.2576843798160553, "learning_rate": 5.0973418854171475e-05, "loss": 0.0249, "step": 41530 }, { "grad_norm": 0.360522598028183, "learning_rate": 5.095275431608447e-05, "loss": 0.0211, "step": 41540 }, { "grad_norm": 0.41000083088874817, "learning_rate": 5.093208961519782e-05, "loss": 0.0225, "step": 41550 }, { "grad_norm": 0.2790673077106476, "learning_rate": 5.091142475504255e-05, "loss": 0.0186, "step": 41560 }, { "grad_norm": 0.436906635761261, "learning_rate": 5.089075973914971e-05, "loss": 0.0243, "step": 41570 }, { "grad_norm": 0.26260945200920105, "learning_rate": 5.08700945710504e-05, "loss": 0.034, "step": 41580 }, { "grad_norm": 0.30959969758987427, "learning_rate": 5.0849429254275714e-05, "loss": 0.0215, "step": 41590 }, { "grad_norm": 0.19494175910949707, "learning_rate": 5.0828763792356804e-05, "loss": 0.0192, "step": 41600 }, { "grad_norm": 0.2416929453611374, "learning_rate": 5.080809818882483e-05, "loss": 0.0246, "step": 41610 }, { "grad_norm": 0.3167884349822998, "learning_rate": 5.078743244721097e-05, "loss": 0.0248, "step": 41620 }, { "grad_norm": 0.25034111738204956, "learning_rate": 5.0766766571046455e-05, "loss": 0.0234, "step": 41630 }, { "grad_norm": 0.45153796672821045, "learning_rate": 5.07461005638625e-05, "loss": 0.0203, "step": 41640 }, { "grad_norm": 0.2153656929731369, "learning_rate": 5.072543442919037e-05, "loss": 0.0246, "step": 41650 }, { "grad_norm": 0.3318409025669098, "learning_rate": 5.070476817056132e-05, "loss": 0.0281, "step": 41660 }, { "grad_norm": 0.2726128399372101, "learning_rate": 5.068410179150668e-05, "loss": 0.0211, "step": 41670 }, { "grad_norm": 0.32104721665382385, "learning_rate": 5.066343529555775e-05, "loss": 0.0214, "step": 41680 }, { "grad_norm": 0.2751721739768982, "learning_rate": 5.064276868624587e-05, "loss": 0.0206, "step": 41690 }, { "grad_norm": 0.348879873752594, "learning_rate": 5.0622101967102396e-05, "loss": 0.0214, "step": 41700 }, { "grad_norm": 0.3550938367843628, "learning_rate": 5.0601435141658714e-05, "loss": 0.021, "step": 41710 }, { "grad_norm": 0.3504195809364319, "learning_rate": 5.058076821344622e-05, "loss": 0.0239, "step": 41720 }, { "grad_norm": 0.48127439618110657, "learning_rate": 5.056010118599632e-05, "loss": 0.0345, "step": 41730 }, { "grad_norm": 0.18561656773090363, "learning_rate": 5.053943406284044e-05, "loss": 0.0217, "step": 41740 }, { "grad_norm": 0.2795518934726715, "learning_rate": 5.051876684751006e-05, "loss": 0.0205, "step": 41750 }, { "grad_norm": 0.3064408004283905, "learning_rate": 5.0498099543536584e-05, "loss": 0.0196, "step": 41760 }, { "grad_norm": 0.2744419574737549, "learning_rate": 5.047743215445152e-05, "loss": 0.0243, "step": 41770 }, { "grad_norm": 0.42652568221092224, "learning_rate": 5.045676468378637e-05, "loss": 0.0235, "step": 41780 }, { "grad_norm": 0.3382469117641449, "learning_rate": 5.0436097135072626e-05, "loss": 0.0282, "step": 41790 }, { "grad_norm": 0.27480876445770264, "learning_rate": 5.041542951184181e-05, "loss": 0.0276, "step": 41800 }, { "grad_norm": 0.2342631220817566, "learning_rate": 5.039476181762545e-05, "loss": 0.0213, "step": 41810 }, { "grad_norm": 0.20279070734977722, "learning_rate": 5.037409405595508e-05, "loss": 0.021, "step": 41820 }, { "grad_norm": 0.3433750867843628, "learning_rate": 5.035342623036229e-05, "loss": 0.0233, "step": 41830 }, { "grad_norm": 0.312678724527359, "learning_rate": 5.033275834437862e-05, "loss": 0.0232, "step": 41840 }, { "grad_norm": 0.3543565273284912, "learning_rate": 5.031209040153564e-05, "loss": 0.0213, "step": 41850 }, { "grad_norm": 0.2971276640892029, "learning_rate": 5.0291422405364955e-05, "loss": 0.0319, "step": 41860 }, { "grad_norm": 0.23459337651729584, "learning_rate": 5.0270754359398133e-05, "loss": 0.0178, "step": 41870 }, { "grad_norm": 0.1810121089220047, "learning_rate": 5.025008626716682e-05, "loss": 0.0169, "step": 41880 }, { "grad_norm": 0.277095764875412, "learning_rate": 5.0229418132202585e-05, "loss": 0.019, "step": 41890 }, { "grad_norm": 0.19489459693431854, "learning_rate": 5.020874995803707e-05, "loss": 0.0252, "step": 41900 }, { "grad_norm": 0.3324931263923645, "learning_rate": 5.01880817482019e-05, "loss": 0.0235, "step": 41910 }, { "grad_norm": 0.41247454285621643, "learning_rate": 5.01674135062287e-05, "loss": 0.0173, "step": 41920 }, { "grad_norm": 0.3197670578956604, "learning_rate": 5.0146745235649115e-05, "loss": 0.0219, "step": 41930 }, { "grad_norm": 0.32390788197517395, "learning_rate": 5.0126076939994795e-05, "loss": 0.0206, "step": 41940 }, { "grad_norm": 0.237025186419487, "learning_rate": 5.010540862279736e-05, "loss": 0.0255, "step": 41950 }, { "grad_norm": 0.20579198002815247, "learning_rate": 5.008474028758846e-05, "loss": 0.0189, "step": 41960 }, { "grad_norm": 0.3047589063644409, "learning_rate": 5.0064071937899784e-05, "loss": 0.022, "step": 41970 }, { "grad_norm": 0.4011688530445099, "learning_rate": 5.004340357726296e-05, "loss": 0.0222, "step": 41980 }, { "grad_norm": 0.24970148503780365, "learning_rate": 5.002273520920965e-05, "loss": 0.0217, "step": 41990 }, { "grad_norm": 0.27705395221710205, "learning_rate": 5.000206683727151e-05, "loss": 0.0181, "step": 42000 }, { "grad_norm": 0.21090815961360931, "learning_rate": 4.998139846498021e-05, "loss": 0.0231, "step": 42010 }, { "grad_norm": 0.33300110697746277, "learning_rate": 4.9960730095867405e-05, "loss": 0.0217, "step": 42020 }, { "grad_norm": 0.18792974948883057, "learning_rate": 4.9940061733464755e-05, "loss": 0.0248, "step": 42030 }, { "grad_norm": 3.550520658493042, "learning_rate": 4.991939338130392e-05, "loss": 0.0295, "step": 42040 }, { "grad_norm": 0.4199523627758026, "learning_rate": 4.989872504291653e-05, "loss": 0.0275, "step": 42050 }, { "grad_norm": 0.21811029314994812, "learning_rate": 4.9878056721834273e-05, "loss": 0.0204, "step": 42060 }, { "grad_norm": 0.3646338880062103, "learning_rate": 4.98573884215888e-05, "loss": 0.0249, "step": 42070 }, { "grad_norm": 0.28932181000709534, "learning_rate": 4.9836720145711715e-05, "loss": 0.0276, "step": 42080 }, { "grad_norm": 0.34595468640327454, "learning_rate": 4.9816051897734725e-05, "loss": 0.0254, "step": 42090 }, { "grad_norm": 0.2746027112007141, "learning_rate": 4.979538368118941e-05, "loss": 0.0241, "step": 42100 }, { "grad_norm": 0.3596470355987549, "learning_rate": 4.9774715499607446e-05, "loss": 0.0218, "step": 42110 }, { "grad_norm": 0.22811207175254822, "learning_rate": 4.9754047356520416e-05, "loss": 0.029, "step": 42120 }, { "grad_norm": 0.2560841739177704, "learning_rate": 4.973337925545999e-05, "loss": 0.0228, "step": 42130 }, { "grad_norm": 0.3326871991157532, "learning_rate": 4.9712711199957746e-05, "loss": 0.0236, "step": 42140 }, { "grad_norm": 0.37493839859962463, "learning_rate": 4.969204319354531e-05, "loss": 0.0267, "step": 42150 }, { "grad_norm": 0.3069152235984802, "learning_rate": 4.9671375239754267e-05, "loss": 0.0214, "step": 42160 }, { "grad_norm": 0.25314798951148987, "learning_rate": 4.9650707342116195e-05, "loss": 0.0203, "step": 42170 }, { "grad_norm": 0.20782999694347382, "learning_rate": 4.96300395041627e-05, "loss": 0.0213, "step": 42180 }, { "grad_norm": 0.3969986140727997, "learning_rate": 4.960937172942532e-05, "loss": 0.0217, "step": 42190 }, { "grad_norm": 0.2450314462184906, "learning_rate": 4.958870402143565e-05, "loss": 0.0198, "step": 42200 }, { "grad_norm": 0.2867638170719147, "learning_rate": 4.9568036383725186e-05, "loss": 0.0213, "step": 42210 }, { "grad_norm": 0.3895511031150818, "learning_rate": 4.9547368819825516e-05, "loss": 0.0263, "step": 42220 }, { "grad_norm": 0.19735674560070038, "learning_rate": 4.952670133326812e-05, "loss": 0.0245, "step": 42230 }, { "grad_norm": 0.27100709080696106, "learning_rate": 4.950603392758453e-05, "loss": 0.0224, "step": 42240 }, { "grad_norm": 0.29487285017967224, "learning_rate": 4.948536660630621e-05, "loss": 0.0221, "step": 42250 }, { "grad_norm": 0.29212459921836853, "learning_rate": 4.9464699372964685e-05, "loss": 0.0237, "step": 42260 }, { "grad_norm": 0.22793784737586975, "learning_rate": 4.9444032231091395e-05, "loss": 0.0251, "step": 42270 }, { "grad_norm": 0.48244062066078186, "learning_rate": 4.9423365184217776e-05, "loss": 0.0189, "step": 42280 }, { "grad_norm": 0.38647884130477905, "learning_rate": 4.940269823587529e-05, "loss": 0.02, "step": 42290 }, { "grad_norm": 0.35115161538124084, "learning_rate": 4.938203138959533e-05, "loss": 0.0174, "step": 42300 }, { "grad_norm": 0.35145825147628784, "learning_rate": 4.9361364648909325e-05, "loss": 0.0251, "step": 42310 }, { "grad_norm": 0.26262426376342773, "learning_rate": 4.9340698017348605e-05, "loss": 0.0176, "step": 42320 }, { "grad_norm": 0.306110143661499, "learning_rate": 4.932003149844458e-05, "loss": 0.0208, "step": 42330 }, { "grad_norm": 0.4684114158153534, "learning_rate": 4.929936509572857e-05, "loss": 0.032, "step": 42340 }, { "grad_norm": 0.40355923771858215, "learning_rate": 4.927869881273191e-05, "loss": 0.0255, "step": 42350 }, { "grad_norm": 0.2792556583881378, "learning_rate": 4.9258032652985894e-05, "loss": 0.0229, "step": 42360 }, { "grad_norm": 0.261282742023468, "learning_rate": 4.9237366620021786e-05, "loss": 0.0219, "step": 42370 }, { "grad_norm": 0.3245570957660675, "learning_rate": 4.921670071737089e-05, "loss": 0.0206, "step": 42380 }, { "grad_norm": 0.37373360991477966, "learning_rate": 4.91960349485644e-05, "loss": 0.0279, "step": 42390 }, { "grad_norm": 0.30899369716644287, "learning_rate": 4.9175369317133566e-05, "loss": 0.0233, "step": 42400 }, { "grad_norm": 0.2671068012714386, "learning_rate": 4.915470382660953e-05, "loss": 0.0234, "step": 42410 }, { "grad_norm": 0.24777843058109283, "learning_rate": 4.9134038480523524e-05, "loss": 0.0213, "step": 42420 }, { "grad_norm": 0.26490989327430725, "learning_rate": 4.911337328240664e-05, "loss": 0.0269, "step": 42430 }, { "grad_norm": 0.2553662061691284, "learning_rate": 4.909270823579003e-05, "loss": 0.0259, "step": 42440 }, { "grad_norm": 0.41984862089157104, "learning_rate": 4.907204334420476e-05, "loss": 0.0203, "step": 42450 }, { "grad_norm": 0.3668266832828522, "learning_rate": 4.9051378611181884e-05, "loss": 0.0266, "step": 42460 }, { "grad_norm": 0.281267911195755, "learning_rate": 4.9030714040252486e-05, "loss": 0.0157, "step": 42470 }, { "grad_norm": 0.21997536718845367, "learning_rate": 4.901004963494752e-05, "loss": 0.0175, "step": 42480 }, { "grad_norm": 0.41087979078292847, "learning_rate": 4.898938539879802e-05, "loss": 0.0266, "step": 42490 }, { "grad_norm": 0.2432912439107895, "learning_rate": 4.8968721335334904e-05, "loss": 0.0251, "step": 42500 }, { "grad_norm": 0.2021198570728302, "learning_rate": 4.894805744808912e-05, "loss": 0.0236, "step": 42510 }, { "grad_norm": 0.2543955147266388, "learning_rate": 4.8927393740591524e-05, "loss": 0.02, "step": 42520 }, { "grad_norm": 0.3774021863937378, "learning_rate": 4.890673021637302e-05, "loss": 0.0211, "step": 42530 }, { "grad_norm": 0.27670276165008545, "learning_rate": 4.888606687896442e-05, "loss": 0.0181, "step": 42540 }, { "grad_norm": 0.38491296768188477, "learning_rate": 4.886540373189652e-05, "loss": 0.0216, "step": 42550 }, { "grad_norm": 0.26251980662345886, "learning_rate": 4.88447407787001e-05, "loss": 0.0146, "step": 42560 }, { "grad_norm": 0.27167195081710815, "learning_rate": 4.8824078022905864e-05, "loss": 0.0178, "step": 42570 }, { "grad_norm": 0.2738015651702881, "learning_rate": 4.880341546804455e-05, "loss": 0.0144, "step": 42580 }, { "grad_norm": 0.3600190281867981, "learning_rate": 4.878275311764679e-05, "loss": 0.019, "step": 42590 }, { "grad_norm": 0.5258971452713013, "learning_rate": 4.8762090975243254e-05, "loss": 0.0203, "step": 42600 }, { "grad_norm": 0.250510573387146, "learning_rate": 4.874142904436448e-05, "loss": 0.0204, "step": 42610 }, { "grad_norm": 0.492378294467926, "learning_rate": 4.8720767328541094e-05, "loss": 0.0324, "step": 42620 }, { "grad_norm": 0.3902287781238556, "learning_rate": 4.870010583130358e-05, "loss": 0.0216, "step": 42630 }, { "grad_norm": 0.3292604386806488, "learning_rate": 4.867944455618241e-05, "loss": 0.0254, "step": 42640 }, { "grad_norm": 0.25476932525634766, "learning_rate": 4.865878350670807e-05, "loss": 0.0184, "step": 42650 }, { "grad_norm": 0.30564191937446594, "learning_rate": 4.8638122686410914e-05, "loss": 0.0195, "step": 42660 }, { "grad_norm": 0.3044644892215729, "learning_rate": 4.861746209882137e-05, "loss": 0.0165, "step": 42670 }, { "grad_norm": 0.2670074701309204, "learning_rate": 4.859680174746972e-05, "loss": 0.0239, "step": 42680 }, { "grad_norm": 0.29079943895339966, "learning_rate": 4.857614163588629e-05, "loss": 0.0207, "step": 42690 }, { "grad_norm": 0.2722053825855255, "learning_rate": 4.855548176760131e-05, "loss": 0.0197, "step": 42700 }, { "grad_norm": 0.27869871258735657, "learning_rate": 4.8534822146144986e-05, "loss": 0.0181, "step": 42710 }, { "grad_norm": 0.35661110281944275, "learning_rate": 4.851416277504749e-05, "loss": 0.0319, "step": 42720 }, { "grad_norm": 0.42964157462120056, "learning_rate": 4.8493503657838923e-05, "loss": 0.017, "step": 42730 }, { "grad_norm": 0.16439978778362274, "learning_rate": 4.84728447980494e-05, "loss": 0.0213, "step": 42740 }, { "grad_norm": 0.43065834045410156, "learning_rate": 4.8452186199208914e-05, "loss": 0.0225, "step": 42750 }, { "grad_norm": 0.3375987112522125, "learning_rate": 4.843152786484749e-05, "loss": 0.0219, "step": 42760 }, { "grad_norm": 0.2760666012763977, "learning_rate": 4.8410869798495035e-05, "loss": 0.0204, "step": 42770 }, { "grad_norm": 0.24622666835784912, "learning_rate": 4.8390212003681486e-05, "loss": 0.0226, "step": 42780 }, { "grad_norm": 0.29836562275886536, "learning_rate": 4.836955448393667e-05, "loss": 0.023, "step": 42790 }, { "grad_norm": 0.26098936796188354, "learning_rate": 4.8348897242790406e-05, "loss": 0.0273, "step": 42800 }, { "grad_norm": 0.20505863428115845, "learning_rate": 4.832824028377243e-05, "loss": 0.0173, "step": 42810 }, { "grad_norm": 0.2834073305130005, "learning_rate": 4.830758361041249e-05, "loss": 0.02, "step": 42820 }, { "grad_norm": 0.42207062244415283, "learning_rate": 4.828692722624022e-05, "loss": 0.0288, "step": 42830 }, { "grad_norm": 0.2565664052963257, "learning_rate": 4.826627113478522e-05, "loss": 0.0308, "step": 42840 }, { "grad_norm": 0.32572048902511597, "learning_rate": 4.824561533957708e-05, "loss": 0.0182, "step": 42850 }, { "grad_norm": 0.3094055950641632, "learning_rate": 4.822495984414527e-05, "loss": 0.0199, "step": 42860 }, { "grad_norm": 0.2161288559436798, "learning_rate": 4.8204304652019304e-05, "loss": 0.0185, "step": 42870 }, { "grad_norm": 0.44879791140556335, "learning_rate": 4.8183649766728537e-05, "loss": 0.0245, "step": 42880 }, { "grad_norm": 0.5073366165161133, "learning_rate": 4.816299519180237e-05, "loss": 0.021, "step": 42890 }, { "grad_norm": 0.20242692530155182, "learning_rate": 4.8142340930770066e-05, "loss": 0.0195, "step": 42900 }, { "grad_norm": 0.2975953221321106, "learning_rate": 4.81216869871609e-05, "loss": 0.0188, "step": 42910 }, { "grad_norm": 0.19849514961242676, "learning_rate": 4.8101033364504064e-05, "loss": 0.0177, "step": 42920 }, { "grad_norm": 0.3617391288280487, "learning_rate": 4.808038006632866e-05, "loss": 0.0243, "step": 42930 }, { "grad_norm": 0.5487425923347473, "learning_rate": 4.805972709616383e-05, "loss": 0.0208, "step": 42940 }, { "grad_norm": 0.5859384536743164, "learning_rate": 4.8039074457538556e-05, "loss": 0.0194, "step": 42950 }, { "grad_norm": 0.2835950255393982, "learning_rate": 4.801842215398184e-05, "loss": 0.0298, "step": 42960 }, { "grad_norm": 0.3248167037963867, "learning_rate": 4.799777018902256e-05, "loss": 0.0249, "step": 42970 }, { "grad_norm": 0.2985663115978241, "learning_rate": 4.797711856618961e-05, "loss": 0.0264, "step": 42980 }, { "grad_norm": 0.44646206498146057, "learning_rate": 4.7956467289011765e-05, "loss": 0.0214, "step": 42990 }, { "grad_norm": 0.30787014961242676, "learning_rate": 4.793581636101778e-05, "loss": 0.0258, "step": 43000 }, { "grad_norm": 0.4801853895187378, "learning_rate": 4.7915165785736326e-05, "loss": 0.0262, "step": 43010 }, { "grad_norm": 0.3961521089076996, "learning_rate": 4.789451556669599e-05, "loss": 0.0195, "step": 43020 }, { "grad_norm": 0.945600152015686, "learning_rate": 4.7873865707425384e-05, "loss": 0.0258, "step": 43030 }, { "grad_norm": 0.3930610716342926, "learning_rate": 4.7853216211452975e-05, "loss": 0.0313, "step": 43040 }, { "grad_norm": 0.3640581965446472, "learning_rate": 4.78325670823072e-05, "loss": 0.0256, "step": 43050 }, { "grad_norm": 0.20152224600315094, "learning_rate": 4.781191832351641e-05, "loss": 0.0273, "step": 43060 }, { "grad_norm": 0.33677375316619873, "learning_rate": 4.7791269938608965e-05, "loss": 0.0263, "step": 43070 }, { "grad_norm": 0.32534241676330566, "learning_rate": 4.777062193111305e-05, "loss": 0.0253, "step": 43080 }, { "grad_norm": 0.4274851381778717, "learning_rate": 4.77499743045569e-05, "loss": 0.0226, "step": 43090 }, { "grad_norm": 0.35679110884666443, "learning_rate": 4.7729327062468596e-05, "loss": 0.0227, "step": 43100 }, { "grad_norm": 0.3103814721107483, "learning_rate": 4.770868020837619e-05, "loss": 0.0197, "step": 43110 }, { "grad_norm": 0.21391718089580536, "learning_rate": 4.768803374580768e-05, "loss": 0.0165, "step": 43120 }, { "grad_norm": 0.2757989764213562, "learning_rate": 4.766738767829094e-05, "loss": 0.0189, "step": 43130 }, { "grad_norm": 0.2582617998123169, "learning_rate": 4.764674200935388e-05, "loss": 0.0227, "step": 43140 }, { "grad_norm": 0.31656184792518616, "learning_rate": 4.762609674252424e-05, "loss": 0.0204, "step": 43150 }, { "grad_norm": 0.22508028149604797, "learning_rate": 4.760545188132974e-05, "loss": 0.034, "step": 43160 }, { "grad_norm": 0.22276464104652405, "learning_rate": 4.758480742929801e-05, "loss": 0.0208, "step": 43170 }, { "grad_norm": 0.3854828476905823, "learning_rate": 4.756416338995664e-05, "loss": 0.0287, "step": 43180 }, { "grad_norm": 0.38968852162361145, "learning_rate": 4.7543519766833145e-05, "loss": 0.0183, "step": 43190 }, { "grad_norm": 0.2539924383163452, "learning_rate": 4.752287656345492e-05, "loss": 0.0258, "step": 43200 }, { "grad_norm": 0.22475118935108185, "learning_rate": 4.7502233783349355e-05, "loss": 0.021, "step": 43210 }, { "grad_norm": 0.3231549859046936, "learning_rate": 4.7481591430043694e-05, "loss": 0.0212, "step": 43220 }, { "grad_norm": 0.2766818404197693, "learning_rate": 4.746094950706521e-05, "loss": 0.025, "step": 43230 }, { "grad_norm": 0.19381695985794067, "learning_rate": 4.744030801794099e-05, "loss": 0.02, "step": 43240 }, { "grad_norm": 0.46676674485206604, "learning_rate": 4.741966696619813e-05, "loss": 0.0305, "step": 43250 }, { "grad_norm": 0.2835930287837982, "learning_rate": 4.739902635536361e-05, "loss": 0.0215, "step": 43260 }, { "grad_norm": 0.32912591099739075, "learning_rate": 4.737838618896436e-05, "loss": 0.0225, "step": 43270 }, { "grad_norm": 0.5077897906303406, "learning_rate": 4.7357746470527203e-05, "loss": 0.0235, "step": 43280 }, { "grad_norm": 0.3644685447216034, "learning_rate": 4.73371072035789e-05, "loss": 0.0237, "step": 43290 }, { "grad_norm": 0.37933576107025146, "learning_rate": 4.731646839164616e-05, "loss": 0.0259, "step": 43300 }, { "grad_norm": 0.303378701210022, "learning_rate": 4.729583003825556e-05, "loss": 0.0242, "step": 43310 }, { "grad_norm": 0.18792782723903656, "learning_rate": 4.727519214693367e-05, "loss": 0.0231, "step": 43320 }, { "grad_norm": 0.2667706310749054, "learning_rate": 4.725455472120689e-05, "loss": 0.0181, "step": 43330 }, { "grad_norm": 0.2010575830936432, "learning_rate": 4.723391776460164e-05, "loss": 0.0225, "step": 43340 }, { "grad_norm": 0.2776174545288086, "learning_rate": 4.7213281280644186e-05, "loss": 0.0224, "step": 43350 }, { "grad_norm": 0.2387990802526474, "learning_rate": 4.719264527286075e-05, "loss": 0.0203, "step": 43360 }, { "grad_norm": 0.18459859490394592, "learning_rate": 4.717200974477744e-05, "loss": 0.0179, "step": 43370 }, { "grad_norm": 0.38000309467315674, "learning_rate": 4.715137469992034e-05, "loss": 0.0227, "step": 43380 }, { "grad_norm": 0.2511669993400574, "learning_rate": 4.713074014181539e-05, "loss": 0.0175, "step": 43390 }, { "grad_norm": 0.4195455014705658, "learning_rate": 4.711010607398846e-05, "loss": 0.0187, "step": 43400 }, { "grad_norm": 0.27758875489234924, "learning_rate": 4.708947249996538e-05, "loss": 0.0195, "step": 43410 }, { "grad_norm": 0.30562084913253784, "learning_rate": 4.706883942327183e-05, "loss": 0.0217, "step": 43420 }, { "grad_norm": 0.2604512572288513, "learning_rate": 4.704820684743347e-05, "loss": 0.0199, "step": 43430 }, { "grad_norm": 0.32432541251182556, "learning_rate": 4.702757477597581e-05, "loss": 0.025, "step": 43440 }, { "grad_norm": 0.30048295855522156, "learning_rate": 4.700694321242435e-05, "loss": 0.0203, "step": 43450 }, { "grad_norm": 0.4179830551147461, "learning_rate": 4.698631216030441e-05, "loss": 0.0203, "step": 43460 }, { "grad_norm": 0.26703017950057983, "learning_rate": 4.6965681623141314e-05, "loss": 0.0199, "step": 43470 }, { "grad_norm": 0.3095860779285431, "learning_rate": 4.694505160446024e-05, "loss": 0.0166, "step": 43480 }, { "grad_norm": 0.22601301968097687, "learning_rate": 4.692442210778628e-05, "loss": 0.0202, "step": 43490 }, { "grad_norm": 0.4215850234031677, "learning_rate": 4.690379313664448e-05, "loss": 0.0213, "step": 43500 }, { "grad_norm": 0.3013453185558319, "learning_rate": 4.688316469455973e-05, "loss": 0.0227, "step": 43510 }, { "grad_norm": 0.22949840128421783, "learning_rate": 4.6862536785056914e-05, "loss": 0.0245, "step": 43520 }, { "grad_norm": 0.284261554479599, "learning_rate": 4.684190941166073e-05, "loss": 0.0222, "step": 43530 }, { "grad_norm": 0.38527169823646545, "learning_rate": 4.682128257789587e-05, "loss": 0.0194, "step": 43540 }, { "grad_norm": 0.3316204845905304, "learning_rate": 4.680065628728687e-05, "loss": 0.0191, "step": 43550 }, { "grad_norm": 0.4596492648124695, "learning_rate": 4.678003054335822e-05, "loss": 0.029, "step": 43560 }, { "grad_norm": 0.3342460095882416, "learning_rate": 4.675940534963428e-05, "loss": 0.0244, "step": 43570 }, { "grad_norm": 0.2732212543487549, "learning_rate": 4.673878070963931e-05, "loss": 0.0221, "step": 43580 }, { "grad_norm": 0.17517858743667603, "learning_rate": 4.671815662689756e-05, "loss": 0.0176, "step": 43590 }, { "grad_norm": 0.25644543766975403, "learning_rate": 4.669753310493306e-05, "loss": 0.0179, "step": 43600 }, { "grad_norm": 0.4008130431175232, "learning_rate": 4.6676910147269845e-05, "loss": 0.0185, "step": 43610 }, { "grad_norm": 0.24686580896377563, "learning_rate": 4.665628775743177e-05, "loss": 0.0188, "step": 43620 }, { "grad_norm": 0.21024464070796967, "learning_rate": 4.66356659389427e-05, "loss": 0.0194, "step": 43630 }, { "grad_norm": 0.3079661726951599, "learning_rate": 4.661504469532629e-05, "loss": 0.0205, "step": 43640 }, { "grad_norm": 0.31765300035476685, "learning_rate": 4.659442403010617e-05, "loss": 0.0187, "step": 43650 }, { "grad_norm": 0.22280572354793549, "learning_rate": 4.6573803946805845e-05, "loss": 0.0161, "step": 43660 }, { "grad_norm": 0.25556010007858276, "learning_rate": 4.6553184448948695e-05, "loss": 0.0187, "step": 43670 }, { "grad_norm": 0.28525862097740173, "learning_rate": 4.653256554005807e-05, "loss": 0.019, "step": 43680 }, { "grad_norm": 0.2530611753463745, "learning_rate": 4.6511947223657145e-05, "loss": 0.0282, "step": 43690 }, { "grad_norm": 0.3086549937725067, "learning_rate": 4.649132950326906e-05, "loss": 0.0258, "step": 43700 }, { "grad_norm": 0.3116607666015625, "learning_rate": 4.647071238241679e-05, "loss": 0.0243, "step": 43710 }, { "grad_norm": 0.3439277410507202, "learning_rate": 4.6450095864623264e-05, "loss": 0.0161, "step": 43720 }, { "grad_norm": 0.3613927364349365, "learning_rate": 4.642947995341123e-05, "loss": 0.0215, "step": 43730 }, { "grad_norm": 0.1712457686662674, "learning_rate": 4.640886465230345e-05, "loss": 0.02, "step": 43740 }, { "grad_norm": 0.3775070607662201, "learning_rate": 4.6388249964822485e-05, "loss": 0.0202, "step": 43750 }, { "grad_norm": 0.29638785123825073, "learning_rate": 4.63676358944908e-05, "loss": 0.0174, "step": 43760 }, { "grad_norm": 0.4863664507865906, "learning_rate": 4.634702244483081e-05, "loss": 0.0284, "step": 43770 }, { "grad_norm": 0.32649365067481995, "learning_rate": 4.632640961936475e-05, "loss": 0.0207, "step": 43780 }, { "grad_norm": 0.2999378740787506, "learning_rate": 4.6305797421614835e-05, "loss": 0.0192, "step": 43790 }, { "grad_norm": 0.29011231660842896, "learning_rate": 4.6285185855103105e-05, "loss": 0.0215, "step": 43800 }, { "grad_norm": 0.2984180152416229, "learning_rate": 4.626457492335151e-05, "loss": 0.0224, "step": 43810 }, { "grad_norm": 0.19797231256961823, "learning_rate": 4.624396462988188e-05, "loss": 0.016, "step": 43820 }, { "grad_norm": 0.2938765287399292, "learning_rate": 4.6223354978215985e-05, "loss": 0.0183, "step": 43830 }, { "grad_norm": 0.38978517055511475, "learning_rate": 4.620274597187544e-05, "loss": 0.024, "step": 43840 }, { "grad_norm": 0.24682575464248657, "learning_rate": 4.6182137614381736e-05, "loss": 0.0233, "step": 43850 }, { "grad_norm": 0.3051426410675049, "learning_rate": 4.616152990925631e-05, "loss": 0.0169, "step": 43860 }, { "grad_norm": 0.25092193484306335, "learning_rate": 4.614092286002041e-05, "loss": 0.0181, "step": 43870 }, { "grad_norm": 0.22386150062084198, "learning_rate": 4.612031647019528e-05, "loss": 0.017, "step": 43880 }, { "grad_norm": 0.21382877230644226, "learning_rate": 4.609971074330193e-05, "loss": 0.0212, "step": 43890 }, { "grad_norm": 0.32309597730636597, "learning_rate": 4.607910568286136e-05, "loss": 0.0256, "step": 43900 }, { "grad_norm": 0.2857847213745117, "learning_rate": 4.605850129239438e-05, "loss": 0.0198, "step": 43910 }, { "grad_norm": 0.28489217162132263, "learning_rate": 4.603789757542174e-05, "loss": 0.0203, "step": 43920 }, { "grad_norm": 0.3842692971229553, "learning_rate": 4.601729453546404e-05, "loss": 0.0212, "step": 43930 }, { "grad_norm": 0.29337677359580994, "learning_rate": 4.599669217604177e-05, "loss": 0.0241, "step": 43940 }, { "grad_norm": 0.37365105748176575, "learning_rate": 4.597609050067532e-05, "loss": 0.0257, "step": 43950 }, { "grad_norm": 0.23569709062576294, "learning_rate": 4.595548951288495e-05, "loss": 0.0243, "step": 43960 }, { "grad_norm": 0.8864210844039917, "learning_rate": 4.593488921619081e-05, "loss": 0.0272, "step": 43970 }, { "grad_norm": 0.4869515597820282, "learning_rate": 4.591428961411289e-05, "loss": 0.0239, "step": 43980 }, { "grad_norm": 0.28326770663261414, "learning_rate": 4.589369071017117e-05, "loss": 0.0247, "step": 43990 }, { "grad_norm": 0.26193907856941223, "learning_rate": 4.587309250788538e-05, "loss": 0.0193, "step": 44000 }, { "grad_norm": 0.2840624451637268, "learning_rate": 4.585249501077522e-05, "loss": 0.0172, "step": 44010 }, { "grad_norm": 0.1393526941537857, "learning_rate": 4.583189822236021e-05, "loss": 0.0232, "step": 44020 }, { "grad_norm": 0.30205056071281433, "learning_rate": 4.5811302146159816e-05, "loss": 0.0196, "step": 44030 }, { "grad_norm": 0.220318004488945, "learning_rate": 4.579070678569332e-05, "loss": 0.0172, "step": 44040 }, { "grad_norm": 0.31620103120803833, "learning_rate": 4.5770112144479904e-05, "loss": 0.015, "step": 44050 }, { "grad_norm": 0.4443932771682739, "learning_rate": 4.5749518226038645e-05, "loss": 0.0168, "step": 44060 }, { "grad_norm": 0.3494744300842285, "learning_rate": 4.572892503388845e-05, "loss": 0.0245, "step": 44070 }, { "grad_norm": 0.2372060865163803, "learning_rate": 4.570833257154817e-05, "loss": 0.0216, "step": 44080 }, { "grad_norm": 0.2839142978191376, "learning_rate": 4.568774084253646e-05, "loss": 0.0258, "step": 44090 }, { "grad_norm": 0.2968863546848297, "learning_rate": 4.566714985037191e-05, "loss": 0.0221, "step": 44100 }, { "grad_norm": 0.3466372787952423, "learning_rate": 4.564655959857295e-05, "loss": 0.0265, "step": 44110 }, { "grad_norm": 0.33307138085365295, "learning_rate": 4.5625970090657893e-05, "loss": 0.0183, "step": 44120 }, { "grad_norm": 0.3834565281867981, "learning_rate": 4.560538133014491e-05, "loss": 0.0224, "step": 44130 }, { "grad_norm": 0.21831068396568298, "learning_rate": 4.5584793320552055e-05, "loss": 0.0144, "step": 44140 }, { "grad_norm": 0.31213927268981934, "learning_rate": 4.556420606539728e-05, "loss": 0.0204, "step": 44150 }, { "grad_norm": 0.18697240948677063, "learning_rate": 4.554361956819836e-05, "loss": 0.0187, "step": 44160 }, { "grad_norm": 0.23813606798648834, "learning_rate": 4.552303383247299e-05, "loss": 0.0202, "step": 44170 }, { "grad_norm": 0.3006693124771118, "learning_rate": 4.5502448861738664e-05, "loss": 0.0214, "step": 44180 }, { "grad_norm": 0.2718413770198822, "learning_rate": 4.548186465951284e-05, "loss": 0.0233, "step": 44190 }, { "grad_norm": 0.3262570798397064, "learning_rate": 4.546128122931277e-05, "loss": 0.0242, "step": 44200 }, { "grad_norm": 0.24833038449287415, "learning_rate": 4.544069857465561e-05, "loss": 0.0222, "step": 44210 }, { "grad_norm": 0.32513657212257385, "learning_rate": 4.542011669905837e-05, "loss": 0.0205, "step": 44220 }, { "grad_norm": 0.28985798358917236, "learning_rate": 4.53995356060379e-05, "loss": 0.0195, "step": 44230 }, { "grad_norm": 0.28070130944252014, "learning_rate": 4.5378955299110994e-05, "loss": 0.0158, "step": 44240 }, { "grad_norm": 0.4585863947868347, "learning_rate": 4.5358375781794233e-05, "loss": 0.0288, "step": 44250 }, { "grad_norm": 0.2406332939863205, "learning_rate": 4.533779705760411e-05, "loss": 0.0209, "step": 44260 }, { "grad_norm": 0.16217076778411865, "learning_rate": 4.5317219130056934e-05, "loss": 0.0223, "step": 44270 }, { "grad_norm": 0.33010444045066833, "learning_rate": 4.5296642002668964e-05, "loss": 0.0253, "step": 44280 }, { "grad_norm": 0.4474712908267975, "learning_rate": 4.5276065678956216e-05, "loss": 0.0217, "step": 44290 }, { "grad_norm": 0.28709617257118225, "learning_rate": 4.525549016243466e-05, "loss": 0.0208, "step": 44300 }, { "grad_norm": 0.36528104543685913, "learning_rate": 4.523491545662008e-05, "loss": 0.0213, "step": 44310 }, { "grad_norm": 0.4172581434249878, "learning_rate": 4.5214341565028104e-05, "loss": 0.0251, "step": 44320 }, { "grad_norm": 0.4220332205295563, "learning_rate": 4.519376849117428e-05, "loss": 0.0158, "step": 44330 }, { "grad_norm": 0.2595018446445465, "learning_rate": 4.517319623857395e-05, "loss": 0.0232, "step": 44340 }, { "grad_norm": 0.31999677419662476, "learning_rate": 4.5152624810742394e-05, "loss": 0.0215, "step": 44350 }, { "grad_norm": 0.3770808279514313, "learning_rate": 4.5132054211194664e-05, "loss": 0.0216, "step": 44360 }, { "grad_norm": 0.3270294666290283, "learning_rate": 4.511148444344574e-05, "loss": 0.0166, "step": 44370 }, { "grad_norm": 0.43253347277641296, "learning_rate": 4.509091551101041e-05, "loss": 0.0295, "step": 44380 }, { "grad_norm": 0.24220557510852814, "learning_rate": 4.507034741740338e-05, "loss": 0.0204, "step": 44390 }, { "grad_norm": 0.35948142409324646, "learning_rate": 4.5049780166139145e-05, "loss": 0.0229, "step": 44400 }, { "grad_norm": 0.33753564953804016, "learning_rate": 4.5029213760732075e-05, "loss": 0.027, "step": 44410 }, { "grad_norm": 0.31843793392181396, "learning_rate": 4.5008648204696434e-05, "loss": 0.0239, "step": 44420 }, { "grad_norm": 0.3030259311199188, "learning_rate": 4.4988083501546284e-05, "loss": 0.0216, "step": 44430 }, { "grad_norm": 0.24699854850769043, "learning_rate": 4.4967519654795606e-05, "loss": 0.0157, "step": 44440 }, { "grad_norm": 0.2576773464679718, "learning_rate": 4.494695666795816e-05, "loss": 0.0212, "step": 44450 }, { "grad_norm": 0.3020165264606476, "learning_rate": 4.4926394544547644e-05, "loss": 0.0293, "step": 44460 }, { "grad_norm": 0.34548625349998474, "learning_rate": 4.490583328807752e-05, "loss": 0.0223, "step": 44470 }, { "grad_norm": 0.18107910454273224, "learning_rate": 4.488527290206117e-05, "loss": 0.0185, "step": 44480 }, { "grad_norm": 0.5889074206352234, "learning_rate": 4.48647133900118e-05, "loss": 0.0195, "step": 44490 }, { "grad_norm": 0.39103609323501587, "learning_rate": 4.484415475544244e-05, "loss": 0.0225, "step": 44500 }, { "grad_norm": 0.27195703983306885, "learning_rate": 4.4823597001866035e-05, "loss": 0.0234, "step": 44510 }, { "grad_norm": 0.31126490235328674, "learning_rate": 4.480304013279532e-05, "loss": 0.0243, "step": 44520 }, { "grad_norm": 0.24812549352645874, "learning_rate": 4.478248415174292e-05, "loss": 0.0203, "step": 44530 }, { "grad_norm": 0.3048727214336395, "learning_rate": 4.476192906222126e-05, "loss": 0.0261, "step": 44540 }, { "grad_norm": 0.3511197566986084, "learning_rate": 4.474137486774268e-05, "loss": 0.0248, "step": 44550 }, { "grad_norm": 0.29662179946899414, "learning_rate": 4.4720821571819296e-05, "loss": 0.0189, "step": 44560 }, { "grad_norm": 0.3070218861103058, "learning_rate": 4.470026917796314e-05, "loss": 0.0309, "step": 44570 }, { "grad_norm": 0.30428963899612427, "learning_rate": 4.4679717689686005e-05, "loss": 0.0411, "step": 44580 }, { "grad_norm": 0.3442229926586151, "learning_rate": 4.465916711049963e-05, "loss": 0.0236, "step": 44590 }, { "grad_norm": 0.3759540021419525, "learning_rate": 4.4638617443915524e-05, "loss": 0.0245, "step": 44600 }, { "grad_norm": 0.42224621772766113, "learning_rate": 4.4618068693445055e-05, "loss": 0.0224, "step": 44610 }, { "grad_norm": 0.360574871301651, "learning_rate": 4.459752086259946e-05, "loss": 0.0245, "step": 44620 }, { "grad_norm": 0.4561319649219513, "learning_rate": 4.457697395488977e-05, "loss": 0.0281, "step": 44630 }, { "grad_norm": 0.22280187904834747, "learning_rate": 4.455642797382693e-05, "loss": 0.036, "step": 44640 }, { "grad_norm": 0.2475467175245285, "learning_rate": 4.4535882922921643e-05, "loss": 0.0262, "step": 44650 }, { "grad_norm": 0.5717828273773193, "learning_rate": 4.451533880568455e-05, "loss": 0.0198, "step": 44660 }, { "grad_norm": 0.4149084687232971, "learning_rate": 4.449479562562603e-05, "loss": 0.0261, "step": 44670 }, { "grad_norm": 0.41155436635017395, "learning_rate": 4.447425338625639e-05, "loss": 0.0229, "step": 44680 }, { "grad_norm": 0.34811580181121826, "learning_rate": 4.4453712091085705e-05, "loss": 0.0238, "step": 44690 }, { "grad_norm": 0.5237314701080322, "learning_rate": 4.443317174362392e-05, "loss": 0.0248, "step": 44700 }, { "grad_norm": 0.4325842559337616, "learning_rate": 4.441263234738085e-05, "loss": 0.0205, "step": 44710 }, { "grad_norm": 0.3026754558086395, "learning_rate": 4.4392093905866086e-05, "loss": 0.0211, "step": 44720 }, { "grad_norm": 0.45074647665023804, "learning_rate": 4.437155642258911e-05, "loss": 0.021, "step": 44730 }, { "grad_norm": 0.41175034642219543, "learning_rate": 4.4351019901059177e-05, "loss": 0.0317, "step": 44740 }, { "grad_norm": 0.3003253936767578, "learning_rate": 4.433048434478547e-05, "loss": 0.0306, "step": 44750 }, { "grad_norm": 0.27202191948890686, "learning_rate": 4.4309949757276906e-05, "loss": 0.0259, "step": 44760 }, { "grad_norm": 0.3402796983718872, "learning_rate": 4.428941614204233e-05, "loss": 0.0284, "step": 44770 }, { "grad_norm": 0.28097593784332275, "learning_rate": 4.426888350259034e-05, "loss": 0.0255, "step": 44780 }, { "grad_norm": 0.2502112090587616, "learning_rate": 4.424835184242939e-05, "loss": 0.0287, "step": 44790 }, { "grad_norm": 0.30645009875297546, "learning_rate": 4.422782116506784e-05, "loss": 0.0207, "step": 44800 }, { "grad_norm": 0.3623367249965668, "learning_rate": 4.4207291474013754e-05, "loss": 0.0231, "step": 44810 }, { "grad_norm": 0.2784683108329773, "learning_rate": 4.418676277277514e-05, "loss": 0.026, "step": 44820 }, { "grad_norm": 0.44725093245506287, "learning_rate": 4.416623506485976e-05, "loss": 0.0287, "step": 44830 }, { "grad_norm": 0.33358827233314514, "learning_rate": 4.414570835377527e-05, "loss": 0.0205, "step": 44840 }, { "grad_norm": 0.19830049574375153, "learning_rate": 4.412518264302909e-05, "loss": 0.018, "step": 44850 }, { "grad_norm": 0.3143663704395294, "learning_rate": 4.4104657936128535e-05, "loss": 0.0184, "step": 44860 }, { "grad_norm": 0.789393961429596, "learning_rate": 4.408413423658071e-05, "loss": 0.0271, "step": 44870 }, { "grad_norm": 0.3532160818576813, "learning_rate": 4.406361154789253e-05, "loss": 0.0202, "step": 44880 }, { "grad_norm": 0.452985554933548, "learning_rate": 4.4043089873570776e-05, "loss": 0.0237, "step": 44890 }, { "grad_norm": 0.39630672335624695, "learning_rate": 4.4022569217122025e-05, "loss": 0.0182, "step": 44900 }, { "grad_norm": 0.38274067640304565, "learning_rate": 4.400204958205274e-05, "loss": 0.0187, "step": 44910 }, { "grad_norm": 0.3967168927192688, "learning_rate": 4.3981530971869125e-05, "loss": 0.03, "step": 44920 }, { "grad_norm": 0.2882329225540161, "learning_rate": 4.396101339007727e-05, "loss": 0.016, "step": 44930 }, { "grad_norm": 0.17431725561618805, "learning_rate": 4.394049684018304e-05, "loss": 0.0197, "step": 44940 }, { "grad_norm": 0.3750459551811218, "learning_rate": 4.3919981325692186e-05, "loss": 0.0172, "step": 44950 }, { "grad_norm": 0.1717016100883484, "learning_rate": 4.389946685011024e-05, "loss": 0.0203, "step": 44960 }, { "grad_norm": 0.32908427715301514, "learning_rate": 4.387895341694255e-05, "loss": 0.0348, "step": 44970 }, { "grad_norm": 0.2306809276342392, "learning_rate": 4.3858441029694324e-05, "loss": 0.0238, "step": 44980 }, { "grad_norm": 0.23660868406295776, "learning_rate": 4.3837929691870527e-05, "loss": 0.022, "step": 44990 }, { "grad_norm": 0.3102591037750244, "learning_rate": 4.381741940697604e-05, "loss": 0.0217, "step": 45000 }, { "grad_norm": 0.21437866985797882, "learning_rate": 4.379691017851547e-05, "loss": 0.0192, "step": 45010 }, { "grad_norm": 0.2704688012599945, "learning_rate": 4.3776402009993304e-05, "loss": 0.0275, "step": 45020 }, { "grad_norm": 0.3405372202396393, "learning_rate": 4.3755894904913794e-05, "loss": 0.0209, "step": 45030 }, { "grad_norm": 0.3316006362438202, "learning_rate": 4.373538886678109e-05, "loss": 0.0176, "step": 45040 }, { "grad_norm": 0.2747145891189575, "learning_rate": 4.371488389909909e-05, "loss": 0.017, "step": 45050 }, { "grad_norm": 0.4817500710487366, "learning_rate": 4.3694380005371515e-05, "loss": 0.0177, "step": 45060 }, { "grad_norm": 0.39682695269584656, "learning_rate": 4.367387718910196e-05, "loss": 0.0223, "step": 45070 }, { "grad_norm": 0.31338271498680115, "learning_rate": 4.3653375453793764e-05, "loss": 0.0169, "step": 45080 }, { "grad_norm": 0.24588318169116974, "learning_rate": 4.3632874802950136e-05, "loss": 0.0209, "step": 45090 }, { "grad_norm": 0.32613494992256165, "learning_rate": 4.3612375240074034e-05, "loss": 0.0209, "step": 45100 }, { "grad_norm": 0.24085363745689392, "learning_rate": 4.3591876768668325e-05, "loss": 0.0201, "step": 45110 }, { "grad_norm": 0.3278424143791199, "learning_rate": 4.3571379392235605e-05, "loss": 0.0173, "step": 45120 }, { "grad_norm": 0.38973596692085266, "learning_rate": 4.3550883114278335e-05, "loss": 0.0264, "step": 45130 }, { "grad_norm": 0.23572519421577454, "learning_rate": 4.353038793829876e-05, "loss": 0.0194, "step": 45140 }, { "grad_norm": 0.3226654827594757, "learning_rate": 4.350989386779891e-05, "loss": 0.0189, "step": 45150 }, { "grad_norm": 0.2894551455974579, "learning_rate": 4.3489400906280724e-05, "loss": 0.0329, "step": 45160 }, { "grad_norm": 0.3176802694797516, "learning_rate": 4.3468909057245845e-05, "loss": 0.0207, "step": 45170 }, { "grad_norm": 0.2928822338581085, "learning_rate": 4.3448418324195794e-05, "loss": 0.025, "step": 45180 }, { "grad_norm": 0.20476822555065155, "learning_rate": 4.342792871063184e-05, "loss": 0.0234, "step": 45190 }, { "grad_norm": 0.25262650847435, "learning_rate": 4.3407440220055145e-05, "loss": 0.0199, "step": 45200 }, { "grad_norm": 0.18167023360729218, "learning_rate": 4.33869528559666e-05, "loss": 0.0209, "step": 45210 }, { "grad_norm": 0.2756645977497101, "learning_rate": 4.336646662186696e-05, "loss": 0.0161, "step": 45220 }, { "grad_norm": 0.22038817405700684, "learning_rate": 4.334598152125672e-05, "loss": 0.0238, "step": 45230 }, { "grad_norm": 0.2571990191936493, "learning_rate": 4.3325497557636276e-05, "loss": 0.0275, "step": 45240 }, { "grad_norm": 0.1664973795413971, "learning_rate": 4.330501473450574e-05, "loss": 0.0224, "step": 45250 }, { "grad_norm": 0.3522739112377167, "learning_rate": 4.328453305536507e-05, "loss": 0.0193, "step": 45260 }, { "grad_norm": 0.22737416625022888, "learning_rate": 4.326405252371404e-05, "loss": 0.0172, "step": 45270 }, { "grad_norm": 0.2823866307735443, "learning_rate": 4.324357314305221e-05, "loss": 0.0225, "step": 45280 }, { "grad_norm": 0.30679506063461304, "learning_rate": 4.3223094916878945e-05, "loss": 0.0159, "step": 45290 }, { "grad_norm": 0.3032590448856354, "learning_rate": 4.320261784869338e-05, "loss": 0.0293, "step": 45300 }, { "grad_norm": 0.1460491269826889, "learning_rate": 4.318214194199455e-05, "loss": 0.0251, "step": 45310 }, { "grad_norm": 0.159894660115242, "learning_rate": 4.316166720028118e-05, "loss": 0.0176, "step": 45320 }, { "grad_norm": 0.30135810375213623, "learning_rate": 4.3141193627051864e-05, "loss": 0.0207, "step": 45330 }, { "grad_norm": 0.39805832505226135, "learning_rate": 4.312072122580496e-05, "loss": 0.0242, "step": 45340 }, { "grad_norm": 0.35118985176086426, "learning_rate": 4.3100250000038646e-05, "loss": 0.025, "step": 45350 }, { "grad_norm": 0.38473623991012573, "learning_rate": 4.307977995325091e-05, "loss": 0.0253, "step": 45360 }, { "grad_norm": 0.29590681195259094, "learning_rate": 4.30593110889395e-05, "loss": 0.0276, "step": 45370 }, { "grad_norm": 0.26406461000442505, "learning_rate": 4.3038843410602016e-05, "loss": 0.0193, "step": 45380 }, { "grad_norm": 0.19856685400009155, "learning_rate": 4.3018376921735774e-05, "loss": 0.0251, "step": 45390 }, { "grad_norm": 0.2199050486087799, "learning_rate": 4.299791162583799e-05, "loss": 0.0164, "step": 45400 }, { "grad_norm": 0.3987211287021637, "learning_rate": 4.29774475264056e-05, "loss": 0.0187, "step": 45410 }, { "grad_norm": 0.30784130096435547, "learning_rate": 4.2956984626935365e-05, "loss": 0.0168, "step": 45420 }, { "grad_norm": 0.14975127577781677, "learning_rate": 4.293652293092383e-05, "loss": 0.0166, "step": 45430 }, { "grad_norm": 0.4565017819404602, "learning_rate": 4.2916062441867324e-05, "loss": 0.0201, "step": 45440 }, { "grad_norm": 0.3104560673236847, "learning_rate": 4.289560316326201e-05, "loss": 0.0224, "step": 45450 }, { "grad_norm": 0.37660086154937744, "learning_rate": 4.28751450986038e-05, "loss": 0.0212, "step": 45460 }, { "grad_norm": 0.4492807388305664, "learning_rate": 4.2854688251388444e-05, "loss": 0.0258, "step": 45470 }, { "grad_norm": 0.36093905568122864, "learning_rate": 4.2834232625111425e-05, "loss": 0.0248, "step": 45480 }, { "grad_norm": 0.2757917642593384, "learning_rate": 4.2813778223268086e-05, "loss": 0.0173, "step": 45490 }, { "grad_norm": 0.2765466272830963, "learning_rate": 4.2793325049353477e-05, "loss": 0.0203, "step": 45500 }, { "grad_norm": 0.22687213122844696, "learning_rate": 4.2772873106862535e-05, "loss": 0.0196, "step": 45510 }, { "grad_norm": 0.336845338344574, "learning_rate": 4.275242239928993e-05, "loss": 0.0221, "step": 45520 }, { "grad_norm": 0.2781645953655243, "learning_rate": 4.273197293013009e-05, "loss": 0.0148, "step": 45530 }, { "grad_norm": 0.18251103162765503, "learning_rate": 4.271152470287731e-05, "loss": 0.0171, "step": 45540 }, { "grad_norm": 0.2817625403404236, "learning_rate": 4.2691077721025594e-05, "loss": 0.0165, "step": 45550 }, { "grad_norm": 0.2991929352283478, "learning_rate": 4.267063198806883e-05, "loss": 0.0196, "step": 45560 }, { "grad_norm": 0.4187549352645874, "learning_rate": 4.2650187507500574e-05, "loss": 0.0152, "step": 45570 }, { "grad_norm": 0.1962500363588333, "learning_rate": 4.2629744282814275e-05, "loss": 0.019, "step": 45580 }, { "grad_norm": 0.2661709785461426, "learning_rate": 4.2609302317503074e-05, "loss": 0.0151, "step": 45590 }, { "grad_norm": 0.34281620383262634, "learning_rate": 4.258886161505999e-05, "loss": 0.0208, "step": 45600 }, { "grad_norm": 0.24646824598312378, "learning_rate": 4.2568422178977775e-05, "loss": 0.0196, "step": 45610 }, { "grad_norm": 0.2889546751976013, "learning_rate": 4.254798401274894e-05, "loss": 0.0274, "step": 45620 }, { "grad_norm": 0.3296516537666321, "learning_rate": 4.252754711986583e-05, "loss": 0.0181, "step": 45630 }, { "grad_norm": 0.29864823818206787, "learning_rate": 4.250711150382052e-05, "loss": 0.0206, "step": 45640 }, { "grad_norm": 0.2604384422302246, "learning_rate": 4.248667716810495e-05, "loss": 0.0208, "step": 45650 }, { "grad_norm": 0.23970577120780945, "learning_rate": 4.246624411621074e-05, "loss": 0.0213, "step": 45660 }, { "grad_norm": 0.35259681940078735, "learning_rate": 4.244581235162938e-05, "loss": 0.0238, "step": 45670 }, { "grad_norm": 0.23043100535869598, "learning_rate": 4.2425381877852075e-05, "loss": 0.0247, "step": 45680 }, { "grad_norm": 0.4415419101715088, "learning_rate": 4.2404952698369856e-05, "loss": 0.0177, "step": 45690 }, { "grad_norm": 0.35353216528892517, "learning_rate": 4.238452481667349e-05, "loss": 0.0217, "step": 45700 }, { "grad_norm": 0.25350257754325867, "learning_rate": 4.2364098236253526e-05, "loss": 0.026, "step": 45710 }, { "grad_norm": 0.24559907615184784, "learning_rate": 4.2343672960600356e-05, "loss": 0.0222, "step": 45720 }, { "grad_norm": 0.2730408310890198, "learning_rate": 4.232324899320406e-05, "loss": 0.0177, "step": 45730 }, { "grad_norm": 0.25380077958106995, "learning_rate": 4.230282633755457e-05, "loss": 0.0217, "step": 45740 }, { "grad_norm": 0.2300349920988083, "learning_rate": 4.2282404997141515e-05, "loss": 0.0207, "step": 45750 }, { "grad_norm": 0.28291893005371094, "learning_rate": 4.2261984975454397e-05, "loss": 0.0192, "step": 45760 }, { "grad_norm": 0.47105586528778076, "learning_rate": 4.224156627598239e-05, "loss": 0.0208, "step": 45770 }, { "grad_norm": 0.2149682492017746, "learning_rate": 4.222114890221453e-05, "loss": 0.0177, "step": 45780 }, { "grad_norm": 0.419704407453537, "learning_rate": 4.2200732857639546e-05, "loss": 0.0238, "step": 45790 }, { "grad_norm": 0.21196125447750092, "learning_rate": 4.2180318145746035e-05, "loss": 0.0164, "step": 45800 }, { "grad_norm": 0.39866361021995544, "learning_rate": 4.215990477002227e-05, "loss": 0.0198, "step": 45810 }, { "grad_norm": 1.1611180305480957, "learning_rate": 4.2139492733956356e-05, "loss": 0.0178, "step": 45820 }, { "grad_norm": 0.27769869565963745, "learning_rate": 4.211908204103615e-05, "loss": 0.0243, "step": 45830 }, { "grad_norm": 0.4184565246105194, "learning_rate": 4.2098672694749265e-05, "loss": 0.0264, "step": 45840 }, { "grad_norm": 0.36954453587532043, "learning_rate": 4.2078264698583133e-05, "loss": 0.0292, "step": 45850 }, { "grad_norm": 0.2750341296195984, "learning_rate": 4.205785805602488e-05, "loss": 0.0195, "step": 45860 }, { "grad_norm": 0.32058319449424744, "learning_rate": 4.203745277056149e-05, "loss": 0.0216, "step": 45870 }, { "grad_norm": 0.20115193724632263, "learning_rate": 4.201704884567964e-05, "loss": 0.0235, "step": 45880 }, { "grad_norm": 0.4001367390155792, "learning_rate": 4.1996646284865816e-05, "loss": 0.0279, "step": 45890 }, { "grad_norm": 0.22147619724273682, "learning_rate": 4.197624509160625e-05, "loss": 0.0215, "step": 45900 }, { "grad_norm": 0.4040505886077881, "learning_rate": 4.195584526938692e-05, "loss": 0.017, "step": 45910 }, { "grad_norm": 0.28722885251045227, "learning_rate": 4.193544682169365e-05, "loss": 0.0187, "step": 45920 }, { "grad_norm": 0.44107523560523987, "learning_rate": 4.1915049752011946e-05, "loss": 0.017, "step": 45930 }, { "grad_norm": 0.20840886235237122, "learning_rate": 4.189465406382712e-05, "loss": 0.0205, "step": 45940 }, { "grad_norm": 0.22651255130767822, "learning_rate": 4.187425976062422e-05, "loss": 0.0155, "step": 45950 }, { "grad_norm": 0.41875630617141724, "learning_rate": 4.18538668458881e-05, "loss": 0.0215, "step": 45960 }, { "grad_norm": 0.25275129079818726, "learning_rate": 4.183347532310333e-05, "loss": 0.0174, "step": 45970 }, { "grad_norm": 0.40441998839378357, "learning_rate": 4.181308519575429e-05, "loss": 0.0207, "step": 45980 }, { "grad_norm": 0.3110315501689911, "learning_rate": 4.179269646732507e-05, "loss": 0.0315, "step": 45990 }, { "grad_norm": 0.15221092104911804, "learning_rate": 4.177230914129954e-05, "loss": 0.0212, "step": 46000 }, { "grad_norm": 0.25061845779418945, "learning_rate": 4.175192322116136e-05, "loss": 0.0228, "step": 46010 }, { "grad_norm": 0.651109516620636, "learning_rate": 4.173153871039391e-05, "loss": 0.0211, "step": 46020 }, { "grad_norm": 0.3865812122821808, "learning_rate": 4.171115561248036e-05, "loss": 0.0225, "step": 46030 }, { "grad_norm": 0.2938998341560364, "learning_rate": 4.16907739309036e-05, "loss": 0.0185, "step": 46040 }, { "grad_norm": 0.2358267605304718, "learning_rate": 4.167039366914633e-05, "loss": 0.0177, "step": 46050 }, { "grad_norm": 0.3638002872467041, "learning_rate": 4.165001483069096e-05, "loss": 0.0217, "step": 46060 }, { "grad_norm": 0.21069413423538208, "learning_rate": 4.16296374190197e-05, "loss": 0.0267, "step": 46070 }, { "grad_norm": 0.33385005593299866, "learning_rate": 4.1609261437614464e-05, "loss": 0.02, "step": 46080 }, { "grad_norm": 0.45254161953926086, "learning_rate": 4.158888688995696e-05, "loss": 0.0244, "step": 46090 }, { "grad_norm": 0.2905113399028778, "learning_rate": 4.1568513779528645e-05, "loss": 0.0171, "step": 46100 }, { "grad_norm": 0.2671397626399994, "learning_rate": 4.1548142109810704e-05, "loss": 0.0282, "step": 46110 }, { "grad_norm": 0.3502446413040161, "learning_rate": 4.152777188428414e-05, "loss": 0.0194, "step": 46120 }, { "grad_norm": 0.20045161247253418, "learning_rate": 4.1507403106429646e-05, "loss": 0.0172, "step": 46130 }, { "grad_norm": 0.29398828744888306, "learning_rate": 4.148703577972768e-05, "loss": 0.0181, "step": 46140 }, { "grad_norm": 0.333676815032959, "learning_rate": 4.146666990765846e-05, "loss": 0.0205, "step": 46150 }, { "grad_norm": 0.27150505781173706, "learning_rate": 4.1446305493701986e-05, "loss": 0.0234, "step": 46160 }, { "grad_norm": 0.3095405101776123, "learning_rate": 4.142594254133796e-05, "loss": 0.0186, "step": 46170 }, { "grad_norm": 0.36980554461479187, "learning_rate": 4.1405581054045835e-05, "loss": 0.0223, "step": 46180 }, { "grad_norm": 0.30875450372695923, "learning_rate": 4.1385221035304864e-05, "loss": 0.0201, "step": 46190 }, { "grad_norm": 0.2877425253391266, "learning_rate": 4.136486248859398e-05, "loss": 0.0228, "step": 46200 }, { "grad_norm": 0.16234751045703888, "learning_rate": 4.1344505417391955e-05, "loss": 0.0229, "step": 46210 }, { "grad_norm": 0.26316505670547485, "learning_rate": 4.132414982517721e-05, "loss": 0.0196, "step": 46220 }, { "grad_norm": 0.20544958114624023, "learning_rate": 4.130379571542798e-05, "loss": 0.0236, "step": 46230 }, { "grad_norm": 0.2267712652683258, "learning_rate": 4.12834430916222e-05, "loss": 0.0188, "step": 46240 }, { "grad_norm": 0.24831756949424744, "learning_rate": 4.126309195723763e-05, "loss": 0.0209, "step": 46250 }, { "grad_norm": 0.25190216302871704, "learning_rate": 4.124274231575168e-05, "loss": 0.0254, "step": 46260 }, { "grad_norm": 0.21952863037586212, "learning_rate": 4.122239417064154e-05, "loss": 0.0197, "step": 46270 }, { "grad_norm": 0.14495877921581268, "learning_rate": 4.1202047525384184e-05, "loss": 0.0144, "step": 46280 }, { "grad_norm": 0.2414568066596985, "learning_rate": 4.118170238345627e-05, "loss": 0.0182, "step": 46290 }, { "grad_norm": 0.17626257240772247, "learning_rate": 4.1161358748334256e-05, "loss": 0.0184, "step": 46300 }, { "grad_norm": 0.28247204422950745, "learning_rate": 4.1141016623494266e-05, "loss": 0.0184, "step": 46310 }, { "grad_norm": 0.31505122780799866, "learning_rate": 4.112067601241227e-05, "loss": 0.0209, "step": 46320 }, { "grad_norm": 0.22639526426792145, "learning_rate": 4.110033691856387e-05, "loss": 0.0181, "step": 46330 }, { "grad_norm": 0.41462358832359314, "learning_rate": 4.107999934542451e-05, "loss": 0.0252, "step": 46340 }, { "grad_norm": 0.19578306376934052, "learning_rate": 4.105966329646928e-05, "loss": 0.0162, "step": 46350 }, { "grad_norm": 0.261348694562912, "learning_rate": 4.103932877517308e-05, "loss": 0.0181, "step": 46360 }, { "grad_norm": 0.256664514541626, "learning_rate": 4.101899578501052e-05, "loss": 0.0181, "step": 46370 }, { "grad_norm": 0.2847203314304352, "learning_rate": 4.099866432945595e-05, "loss": 0.0258, "step": 46380 }, { "grad_norm": 0.3500353693962097, "learning_rate": 4.097833441198346e-05, "loss": 0.0194, "step": 46390 }, { "grad_norm": 0.21518267691135406, "learning_rate": 4.095800603606685e-05, "loss": 0.0187, "step": 46400 }, { "grad_norm": 0.30334246158599854, "learning_rate": 4.093767920517975e-05, "loss": 0.0224, "step": 46410 }, { "grad_norm": 0.26036494970321655, "learning_rate": 4.091735392279539e-05, "loss": 0.0181, "step": 46420 }, { "grad_norm": 0.1920269876718521, "learning_rate": 4.089703019238685e-05, "loss": 0.0193, "step": 46430 }, { "grad_norm": 0.31770646572113037, "learning_rate": 4.0876708017426866e-05, "loss": 0.0203, "step": 46440 }, { "grad_norm": 0.3074573278427124, "learning_rate": 4.085638740138798e-05, "loss": 0.0182, "step": 46450 }, { "grad_norm": 0.3892970383167267, "learning_rate": 4.08360683477424e-05, "loss": 0.0245, "step": 46460 }, { "grad_norm": 0.2611043453216553, "learning_rate": 4.0815750859962085e-05, "loss": 0.0207, "step": 46470 }, { "grad_norm": 0.2703854739665985, "learning_rate": 4.079543494151879e-05, "loss": 0.0244, "step": 46480 }, { "grad_norm": 0.2781912088394165, "learning_rate": 4.07751205958839e-05, "loss": 0.0156, "step": 46490 }, { "grad_norm": 0.33102869987487793, "learning_rate": 4.0754807826528615e-05, "loss": 0.0166, "step": 46500 }, { "grad_norm": 0.23120546340942383, "learning_rate": 4.073449663692379e-05, "loss": 0.0242, "step": 46510 }, { "grad_norm": 0.179197758436203, "learning_rate": 4.07141870305401e-05, "loss": 0.0167, "step": 46520 }, { "grad_norm": 0.48437848687171936, "learning_rate": 4.0693879010847866e-05, "loss": 0.0378, "step": 46530 }, { "grad_norm": 0.2717060148715973, "learning_rate": 4.06735725813172e-05, "loss": 0.0184, "step": 46540 }, { "grad_norm": 0.2653871178627014, "learning_rate": 4.0653267745417903e-05, "loss": 0.0159, "step": 46550 }, { "grad_norm": 0.1552293598651886, "learning_rate": 4.063296450661949e-05, "loss": 0.0136, "step": 46560 }, { "grad_norm": 0.17947833240032196, "learning_rate": 4.061266286839128e-05, "loss": 0.0213, "step": 46570 }, { "grad_norm": 0.5128002166748047, "learning_rate": 4.0592362834202225e-05, "loss": 0.0196, "step": 46580 }, { "grad_norm": 0.19711732864379883, "learning_rate": 4.057206440752107e-05, "loss": 0.0135, "step": 46590 }, { "grad_norm": 0.21990624070167542, "learning_rate": 4.0551767591816245e-05, "loss": 0.0163, "step": 46600 }, { "grad_norm": 0.21783462166786194, "learning_rate": 4.0531472390555935e-05, "loss": 0.0215, "step": 46610 }, { "grad_norm": 0.4009816646575928, "learning_rate": 4.051117880720802e-05, "loss": 0.0194, "step": 46620 }, { "grad_norm": 0.2657136917114258, "learning_rate": 4.049088684524015e-05, "loss": 0.0192, "step": 46630 }, { "grad_norm": 0.2349267601966858, "learning_rate": 4.0470596508119636e-05, "loss": 0.0201, "step": 46640 }, { "grad_norm": 0.19491590559482574, "learning_rate": 4.0450307799313524e-05, "loss": 0.0212, "step": 46650 }, { "grad_norm": 0.285368412733078, "learning_rate": 4.0430020722288656e-05, "loss": 0.0177, "step": 46660 }, { "grad_norm": 0.23340432345867157, "learning_rate": 4.040973528051148e-05, "loss": 0.0209, "step": 46670 }, { "grad_norm": 0.2530601918697357, "learning_rate": 4.038945147744827e-05, "loss": 0.0221, "step": 46680 }, { "grad_norm": 0.2972393035888672, "learning_rate": 4.0369169316564945e-05, "loss": 0.0215, "step": 46690 }, { "grad_norm": 0.229487344622612, "learning_rate": 4.034888880132718e-05, "loss": 0.0209, "step": 46700 }, { "grad_norm": 0.3762955069541931, "learning_rate": 4.032860993520035e-05, "loss": 0.0294, "step": 46710 }, { "grad_norm": 0.30886152386665344, "learning_rate": 4.030833272164959e-05, "loss": 0.0208, "step": 46720 }, { "grad_norm": 0.28146952390670776, "learning_rate": 4.0288057164139705e-05, "loss": 0.027, "step": 46730 }, { "grad_norm": 0.191872239112854, "learning_rate": 4.0267783266135205e-05, "loss": 0.0204, "step": 46740 }, { "grad_norm": 0.21312613785266876, "learning_rate": 4.024751103110039e-05, "loss": 0.0213, "step": 46750 }, { "grad_norm": 0.27400627732276917, "learning_rate": 4.0227240462499176e-05, "loss": 0.026, "step": 46760 }, { "grad_norm": 0.28710469603538513, "learning_rate": 4.020697156379531e-05, "loss": 0.0205, "step": 46770 }, { "grad_norm": 0.34658777713775635, "learning_rate": 4.018670433845215e-05, "loss": 0.0233, "step": 46780 }, { "grad_norm": 0.29957571625709534, "learning_rate": 4.016643878993284e-05, "loss": 0.0206, "step": 46790 }, { "grad_norm": 0.26259323954582214, "learning_rate": 4.014617492170017e-05, "loss": 0.0244, "step": 46800 }, { "grad_norm": 0.2931559681892395, "learning_rate": 4.0125912737216726e-05, "loss": 0.0294, "step": 46810 }, { "grad_norm": 0.2368561327457428, "learning_rate": 4.0105652239944735e-05, "loss": 0.0265, "step": 46820 }, { "grad_norm": 0.19975310564041138, "learning_rate": 4.0085393433346144e-05, "loss": 0.0221, "step": 46830 }, { "grad_norm": 0.6647598743438721, "learning_rate": 4.006513632088268e-05, "loss": 0.018, "step": 46840 }, { "grad_norm": 0.2986154854297638, "learning_rate": 4.004488090601567e-05, "loss": 0.0163, "step": 46850 }, { "grad_norm": 0.23873533308506012, "learning_rate": 4.002462719220626e-05, "loss": 0.0172, "step": 46860 }, { "grad_norm": 0.24022793769836426, "learning_rate": 4.000437518291522e-05, "loss": 0.0229, "step": 46870 }, { "grad_norm": 0.22183136641979218, "learning_rate": 3.9984124881603094e-05, "loss": 0.0197, "step": 46880 }, { "grad_norm": 0.37664443254470825, "learning_rate": 3.9963876291730086e-05, "loss": 0.0198, "step": 46890 }, { "grad_norm": 0.33591228723526, "learning_rate": 3.994362941675614e-05, "loss": 0.0229, "step": 46900 }, { "grad_norm": 0.38794615864753723, "learning_rate": 3.992338426014088e-05, "loss": 0.0188, "step": 46910 }, { "grad_norm": 0.34223440289497375, "learning_rate": 3.9903140825343636e-05, "loss": 0.0183, "step": 46920 }, { "grad_norm": 0.1817292422056198, "learning_rate": 3.98828991158235e-05, "loss": 0.0155, "step": 46930 }, { "grad_norm": 0.3472632169723511, "learning_rate": 3.9862659135039185e-05, "loss": 0.0258, "step": 46940 }, { "grad_norm": 0.2844812572002411, "learning_rate": 3.984242088644918e-05, "loss": 0.0203, "step": 46950 }, { "grad_norm": 0.19023549556732178, "learning_rate": 3.9822184373511615e-05, "loss": 0.0154, "step": 46960 }, { "grad_norm": 0.3455438017845154, "learning_rate": 3.98019495996844e-05, "loss": 0.0219, "step": 46970 }, { "grad_norm": 0.31854957342147827, "learning_rate": 3.978171656842507e-05, "loss": 0.0199, "step": 46980 }, { "grad_norm": 0.38130131363868713, "learning_rate": 3.976148528319091e-05, "loss": 0.0178, "step": 46990 }, { "grad_norm": 0.1830313503742218, "learning_rate": 3.974125574743888e-05, "loss": 0.0208, "step": 47000 }, { "grad_norm": 0.3619287312030792, "learning_rate": 3.9721027964625686e-05, "loss": 0.0205, "step": 47010 }, { "grad_norm": 0.38606634736061096, "learning_rate": 3.9700801938207676e-05, "loss": 0.0186, "step": 47020 }, { "grad_norm": 0.27503275871276855, "learning_rate": 3.9680577671640916e-05, "loss": 0.0256, "step": 47030 }, { "grad_norm": 0.2458455115556717, "learning_rate": 3.966035516838121e-05, "loss": 0.0261, "step": 47040 }, { "grad_norm": 0.20241278409957886, "learning_rate": 3.9640134431884014e-05, "loss": 0.0199, "step": 47050 }, { "grad_norm": 0.17853505909442902, "learning_rate": 3.961991546560451e-05, "loss": 0.0162, "step": 47060 }, { "grad_norm": 0.19239220023155212, "learning_rate": 3.959969827299753e-05, "loss": 0.0201, "step": 47070 }, { "grad_norm": 0.36537161469459534, "learning_rate": 3.9579482857517684e-05, "loss": 0.0215, "step": 47080 }, { "grad_norm": 0.26215752959251404, "learning_rate": 3.955926922261921e-05, "loss": 0.0181, "step": 47090 }, { "grad_norm": 0.1898902952671051, "learning_rate": 3.9539057371756084e-05, "loss": 0.0197, "step": 47100 }, { "grad_norm": 0.28543707728385925, "learning_rate": 3.951884730838195e-05, "loss": 0.0174, "step": 47110 }, { "grad_norm": 0.3681146502494812, "learning_rate": 3.949863903595012e-05, "loss": 0.0178, "step": 47120 }, { "grad_norm": 0.2694316804409027, "learning_rate": 3.947843255791369e-05, "loss": 0.0216, "step": 47130 }, { "grad_norm": 0.4927993714809418, "learning_rate": 3.9458227877725364e-05, "loss": 0.0264, "step": 47140 }, { "grad_norm": 0.29845935106277466, "learning_rate": 3.943802499883758e-05, "loss": 0.0167, "step": 47150 }, { "grad_norm": 0.28076741099357605, "learning_rate": 3.9417823924702437e-05, "loss": 0.0227, "step": 47160 }, { "grad_norm": 0.3246201276779175, "learning_rate": 3.939762465877178e-05, "loss": 0.0215, "step": 47170 }, { "grad_norm": 0.29956120252609253, "learning_rate": 3.937742720449708e-05, "loss": 0.0207, "step": 47180 }, { "grad_norm": 0.22403663396835327, "learning_rate": 3.9357231565329563e-05, "loss": 0.0192, "step": 47190 }, { "grad_norm": 0.27129074931144714, "learning_rate": 3.933703774472008e-05, "loss": 0.0148, "step": 47200 }, { "grad_norm": 0.2645421028137207, "learning_rate": 3.93168457461192e-05, "loss": 0.0162, "step": 47210 }, { "grad_norm": 0.24400773644447327, "learning_rate": 3.9296655572977216e-05, "loss": 0.02, "step": 47220 }, { "grad_norm": 0.31992989778518677, "learning_rate": 3.927646722874404e-05, "loss": 0.0231, "step": 47230 }, { "grad_norm": 0.25456899404525757, "learning_rate": 3.925628071686934e-05, "loss": 0.0212, "step": 47240 }, { "grad_norm": 1.4949930906295776, "learning_rate": 3.9236096040802415e-05, "loss": 0.0256, "step": 47250 }, { "grad_norm": 0.5251023173332214, "learning_rate": 3.9215913203992294e-05, "loss": 0.0271, "step": 47260 }, { "grad_norm": 0.3084522783756256, "learning_rate": 3.9195732209887645e-05, "loss": 0.0207, "step": 47270 }, { "grad_norm": 0.23379866778850555, "learning_rate": 3.9175553061936875e-05, "loss": 0.0195, "step": 47280 }, { "grad_norm": 0.25333136320114136, "learning_rate": 3.9155375763588045e-05, "loss": 0.0264, "step": 47290 }, { "grad_norm": 0.2627795338630676, "learning_rate": 3.913520031828889e-05, "loss": 0.0186, "step": 47300 }, { "grad_norm": 0.2997283339500427, "learning_rate": 3.911502672948685e-05, "loss": 0.0171, "step": 47310 }, { "grad_norm": 0.2646443247795105, "learning_rate": 3.9094855000629014e-05, "loss": 0.0192, "step": 47320 }, { "grad_norm": 0.19222846627235413, "learning_rate": 3.907468513516223e-05, "loss": 0.0194, "step": 47330 }, { "grad_norm": 0.31521958112716675, "learning_rate": 3.905451713653294e-05, "loss": 0.0178, "step": 47340 }, { "grad_norm": 0.3509475290775299, "learning_rate": 3.903435100818731e-05, "loss": 0.0198, "step": 47350 }, { "grad_norm": 0.3639274835586548, "learning_rate": 3.901418675357117e-05, "loss": 0.0215, "step": 47360 }, { "grad_norm": 0.29112866520881653, "learning_rate": 3.8994024376130075e-05, "loss": 0.0276, "step": 47370 }, { "grad_norm": 0.3177427053451538, "learning_rate": 3.8973863879309194e-05, "loss": 0.0189, "step": 47380 }, { "grad_norm": 0.3949465751647949, "learning_rate": 3.8953705266553394e-05, "loss": 0.0191, "step": 47390 }, { "grad_norm": 0.24054637551307678, "learning_rate": 3.893354854130727e-05, "loss": 0.0244, "step": 47400 }, { "grad_norm": 0.31422412395477295, "learning_rate": 3.8913393707015006e-05, "loss": 0.0157, "step": 47410 }, { "grad_norm": 0.44838619232177734, "learning_rate": 3.889324076712056e-05, "loss": 0.026, "step": 47420 }, { "grad_norm": 0.32808029651641846, "learning_rate": 3.8873089725067476e-05, "loss": 0.0204, "step": 47430 }, { "grad_norm": 0.4204249978065491, "learning_rate": 3.885294058429905e-05, "loss": 0.0311, "step": 47440 }, { "grad_norm": 0.30786240100860596, "learning_rate": 3.8832793348258206e-05, "loss": 0.0202, "step": 47450 }, { "grad_norm": 0.27085572481155396, "learning_rate": 3.881264802038756e-05, "loss": 0.0297, "step": 47460 }, { "grad_norm": 0.4180642366409302, "learning_rate": 3.879250460412939e-05, "loss": 0.0282, "step": 47470 }, { "grad_norm": 0.2529999911785126, "learning_rate": 3.8772363102925644e-05, "loss": 0.0164, "step": 47480 }, { "grad_norm": 0.1916564702987671, "learning_rate": 3.875222352021798e-05, "loss": 0.017, "step": 47490 }, { "grad_norm": 0.2852712869644165, "learning_rate": 3.8732085859447686e-05, "loss": 0.0222, "step": 47500 }, { "grad_norm": 0.3639666438102722, "learning_rate": 3.871195012405575e-05, "loss": 0.0173, "step": 47510 }, { "grad_norm": 0.33639952540397644, "learning_rate": 3.869181631748278e-05, "loss": 0.0214, "step": 47520 }, { "grad_norm": 0.3459853529930115, "learning_rate": 3.867168444316915e-05, "loss": 0.0177, "step": 47530 }, { "grad_norm": 0.3491172194480896, "learning_rate": 3.865155450455481e-05, "loss": 0.0198, "step": 47540 }, { "grad_norm": 0.2842698097229004, "learning_rate": 3.8631426505079426e-05, "loss": 0.0206, "step": 47550 }, { "grad_norm": 0.3646044135093689, "learning_rate": 3.8611300448182304e-05, "loss": 0.0318, "step": 47560 }, { "grad_norm": 0.30765703320503235, "learning_rate": 3.859117633730248e-05, "loss": 0.0177, "step": 47570 }, { "grad_norm": 0.3574672043323517, "learning_rate": 3.857105417587858e-05, "loss": 0.0248, "step": 47580 }, { "grad_norm": 0.37751880288124084, "learning_rate": 3.855093396734894e-05, "loss": 0.0179, "step": 47590 }, { "grad_norm": 0.21813862025737762, "learning_rate": 3.8530815715151545e-05, "loss": 0.0296, "step": 47600 }, { "grad_norm": 0.32311004400253296, "learning_rate": 3.851069942272405e-05, "loss": 0.0191, "step": 47610 }, { "grad_norm": 0.23702490329742432, "learning_rate": 3.849058509350382e-05, "loss": 0.021, "step": 47620 }, { "grad_norm": 0.38990387320518494, "learning_rate": 3.8470472730927783e-05, "loss": 0.0306, "step": 47630 }, { "grad_norm": 0.38714125752449036, "learning_rate": 3.845036233843264e-05, "loss": 0.0159, "step": 47640 }, { "grad_norm": 0.2699051797389984, "learning_rate": 3.843025391945469e-05, "loss": 0.0177, "step": 47650 }, { "grad_norm": 0.3329211473464966, "learning_rate": 3.841014747742992e-05, "loss": 0.0188, "step": 47660 }, { "grad_norm": 0.23241457343101501, "learning_rate": 3.839004301579397e-05, "loss": 0.0163, "step": 47670 }, { "grad_norm": 0.44335445761680603, "learning_rate": 3.8369940537982097e-05, "loss": 0.0211, "step": 47680 }, { "grad_norm": 0.23011265695095062, "learning_rate": 3.834984004742933e-05, "loss": 0.0204, "step": 47690 }, { "grad_norm": 0.19937703013420105, "learning_rate": 3.832974154757026e-05, "loss": 0.0153, "step": 47700 }, { "grad_norm": 0.31963732838630676, "learning_rate": 3.830964504183919e-05, "loss": 0.0186, "step": 47710 }, { "grad_norm": 0.2837464511394501, "learning_rate": 3.828955053367003e-05, "loss": 0.0215, "step": 47720 }, { "grad_norm": 0.344072550535202, "learning_rate": 3.826945802649642e-05, "loss": 0.0177, "step": 47730 }, { "grad_norm": 0.17674008011817932, "learning_rate": 3.824936752375159e-05, "loss": 0.0301, "step": 47740 }, { "grad_norm": 0.27539485692977905, "learning_rate": 3.822927902886848e-05, "loss": 0.0206, "step": 47750 }, { "grad_norm": 0.23441149294376373, "learning_rate": 3.8209192545279653e-05, "loss": 0.0157, "step": 47760 }, { "grad_norm": 0.1715942919254303, "learning_rate": 3.8189108076417326e-05, "loss": 0.0165, "step": 47770 }, { "grad_norm": 0.29909655451774597, "learning_rate": 3.816902562571342e-05, "loss": 0.0266, "step": 47780 }, { "grad_norm": 0.23264364898204803, "learning_rate": 3.814894519659944e-05, "loss": 0.0185, "step": 47790 }, { "grad_norm": 0.2578847110271454, "learning_rate": 3.812886679250661e-05, "loss": 0.0166, "step": 47800 }, { "grad_norm": 0.21330222487449646, "learning_rate": 3.810879041686575e-05, "loss": 0.0136, "step": 47810 }, { "grad_norm": 0.49682316184043884, "learning_rate": 3.808871607310741e-05, "loss": 0.0219, "step": 47820 }, { "grad_norm": 0.3876660168170929, "learning_rate": 3.806864376466169e-05, "loss": 0.0197, "step": 47830 }, { "grad_norm": 0.24039195477962494, "learning_rate": 3.804857349495845e-05, "loss": 0.0195, "step": 47840 }, { "grad_norm": 0.5572479963302612, "learning_rate": 3.802850526742713e-05, "loss": 0.0316, "step": 47850 }, { "grad_norm": 0.3003167510032654, "learning_rate": 3.800843908549683e-05, "loss": 0.0159, "step": 47860 }, { "grad_norm": 0.22784683108329773, "learning_rate": 3.7988374952596325e-05, "loss": 0.0173, "step": 47870 }, { "grad_norm": 0.3113084137439728, "learning_rate": 3.7968312872154e-05, "loss": 0.0209, "step": 47880 }, { "grad_norm": 0.31762564182281494, "learning_rate": 3.7948252847597965e-05, "loss": 0.022, "step": 47890 }, { "grad_norm": 0.29827407002449036, "learning_rate": 3.7928194882355885e-05, "loss": 0.0165, "step": 47900 }, { "grad_norm": 0.33401626348495483, "learning_rate": 3.790813897985515e-05, "loss": 0.0212, "step": 47910 }, { "grad_norm": 0.338615745306015, "learning_rate": 3.7888085143522726e-05, "loss": 0.0241, "step": 47920 }, { "grad_norm": 0.39016565680503845, "learning_rate": 3.7868033376785314e-05, "loss": 0.0218, "step": 47930 }, { "grad_norm": 0.27073734998703003, "learning_rate": 3.784798368306919e-05, "loss": 0.0168, "step": 47940 }, { "grad_norm": 0.21253538131713867, "learning_rate": 3.782793606580029e-05, "loss": 0.0161, "step": 47950 }, { "grad_norm": 0.17305919528007507, "learning_rate": 3.7807890528404205e-05, "loss": 0.0177, "step": 47960 }, { "grad_norm": 0.22756057977676392, "learning_rate": 3.778784707430616e-05, "loss": 0.0208, "step": 47970 }, { "grad_norm": 0.21534673869609833, "learning_rate": 3.776780570693107e-05, "loss": 0.015, "step": 47980 }, { "grad_norm": 0.15764009952545166, "learning_rate": 3.774776642970342e-05, "loss": 0.022, "step": 47990 }, { "grad_norm": 0.196132630109787, "learning_rate": 3.77277292460474e-05, "loss": 0.0176, "step": 48000 }, { "grad_norm": 0.266791433095932, "learning_rate": 3.770769415938678e-05, "loss": 0.0148, "step": 48010 }, { "grad_norm": 0.31940850615501404, "learning_rate": 3.768766117314506e-05, "loss": 0.028, "step": 48020 }, { "grad_norm": 0.2693251669406891, "learning_rate": 3.76676302907453e-05, "loss": 0.016, "step": 48030 }, { "grad_norm": 0.2588450610637665, "learning_rate": 3.764760151561021e-05, "loss": 0.0216, "step": 48040 }, { "grad_norm": 0.1956019103527069, "learning_rate": 3.76275748511622e-05, "loss": 0.0247, "step": 48050 }, { "grad_norm": 0.18036086857318878, "learning_rate": 3.7607550300823255e-05, "loss": 0.0191, "step": 48060 }, { "grad_norm": 0.4961787164211273, "learning_rate": 3.7587527868015044e-05, "loss": 0.0372, "step": 48070 }, { "grad_norm": 0.34674733877182007, "learning_rate": 3.756750755615881e-05, "loss": 0.0256, "step": 48080 }, { "grad_norm": 0.3666747212409973, "learning_rate": 3.754748936867553e-05, "loss": 0.0219, "step": 48090 }, { "grad_norm": 0.2516321837902069, "learning_rate": 3.752747330898573e-05, "loss": 0.0208, "step": 48100 }, { "grad_norm": 0.20972943305969238, "learning_rate": 3.750745938050962e-05, "loss": 0.0192, "step": 48110 }, { "grad_norm": 0.24275679886341095, "learning_rate": 3.7487447586667025e-05, "loss": 0.0203, "step": 48120 }, { "grad_norm": 0.33128586411476135, "learning_rate": 3.7467437930877426e-05, "loss": 0.0184, "step": 48130 }, { "grad_norm": 0.2740909159183502, "learning_rate": 3.744743041655992e-05, "loss": 0.0177, "step": 48140 }, { "grad_norm": 0.3066111207008362, "learning_rate": 3.742742504713324e-05, "loss": 0.0192, "step": 48150 }, { "grad_norm": 0.28120100498199463, "learning_rate": 3.740742182601576e-05, "loss": 0.0186, "step": 48160 }, { "grad_norm": 0.48434871435165405, "learning_rate": 3.7387420756625464e-05, "loss": 0.0229, "step": 48170 }, { "grad_norm": 0.40785467624664307, "learning_rate": 3.736742184238002e-05, "loss": 0.0198, "step": 48180 }, { "grad_norm": 0.20975124835968018, "learning_rate": 3.7347425086696684e-05, "loss": 0.0231, "step": 48190 }, { "grad_norm": 0.35182470083236694, "learning_rate": 3.732743049299235e-05, "loss": 0.0263, "step": 48200 }, { "grad_norm": 0.4391389787197113, "learning_rate": 3.730743806468354e-05, "loss": 0.0165, "step": 48210 }, { "grad_norm": 0.1931980401277542, "learning_rate": 3.7287447805186436e-05, "loss": 0.015, "step": 48220 }, { "grad_norm": 0.5019983649253845, "learning_rate": 3.726745971791682e-05, "loss": 0.0217, "step": 48230 }, { "grad_norm": 0.6126194000244141, "learning_rate": 3.724747380629008e-05, "loss": 0.032, "step": 48240 }, { "grad_norm": 0.3111826479434967, "learning_rate": 3.72274900737213e-05, "loss": 0.0213, "step": 48250 }, { "grad_norm": 0.27900412678718567, "learning_rate": 3.7207508523625123e-05, "loss": 0.0197, "step": 48260 }, { "grad_norm": 0.18755535781383514, "learning_rate": 3.718752915941588e-05, "loss": 0.0164, "step": 48270 }, { "grad_norm": 0.37442493438720703, "learning_rate": 3.7167551984507464e-05, "loss": 0.0193, "step": 48280 }, { "grad_norm": 0.3511691987514496, "learning_rate": 3.714757700231346e-05, "loss": 0.0166, "step": 48290 }, { "grad_norm": 0.256623238325119, "learning_rate": 3.712760421624703e-05, "loss": 0.0197, "step": 48300 }, { "grad_norm": 0.288404256105423, "learning_rate": 3.710763362972099e-05, "loss": 0.0186, "step": 48310 }, { "grad_norm": 0.23247888684272766, "learning_rate": 3.708766524614774e-05, "loss": 0.0215, "step": 48320 }, { "grad_norm": 0.22825616598129272, "learning_rate": 3.7067699068939335e-05, "loss": 0.0141, "step": 48330 }, { "grad_norm": 0.17716360092163086, "learning_rate": 3.704773510150748e-05, "loss": 0.0184, "step": 48340 }, { "grad_norm": 0.3613167107105255, "learning_rate": 3.702777334726344e-05, "loss": 0.0229, "step": 48350 }, { "grad_norm": 0.13944824039936066, "learning_rate": 3.7007813809618164e-05, "loss": 0.0128, "step": 48360 }, { "grad_norm": 0.35905522108078003, "learning_rate": 3.6987856491982145e-05, "loss": 0.0248, "step": 48370 }, { "grad_norm": 0.2526610791683197, "learning_rate": 3.696790139776558e-05, "loss": 0.0216, "step": 48380 }, { "grad_norm": 0.2944944500923157, "learning_rate": 3.6947948530378235e-05, "loss": 0.0192, "step": 48390 }, { "grad_norm": 0.2503330111503601, "learning_rate": 3.6927997893229516e-05, "loss": 0.0185, "step": 48400 }, { "grad_norm": 0.2995924949645996, "learning_rate": 3.690804948972845e-05, "loss": 0.0176, "step": 48410 }, { "grad_norm": 0.2380426824092865, "learning_rate": 3.688810332328363e-05, "loss": 0.0198, "step": 48420 }, { "grad_norm": 0.217475026845932, "learning_rate": 3.686815939730336e-05, "loss": 0.0251, "step": 48430 }, { "grad_norm": 0.20553191006183624, "learning_rate": 3.684821771519548e-05, "loss": 0.017, "step": 48440 }, { "grad_norm": 0.28946903347969055, "learning_rate": 3.68282782803675e-05, "loss": 0.0181, "step": 48450 }, { "grad_norm": 0.1591491550207138, "learning_rate": 3.6808341096226504e-05, "loss": 0.0226, "step": 48460 }, { "grad_norm": 0.38462433218955994, "learning_rate": 3.678840616617924e-05, "loss": 0.0192, "step": 48470 }, { "grad_norm": 0.17702758312225342, "learning_rate": 3.6768473493632e-05, "loss": 0.0159, "step": 48480 }, { "grad_norm": 0.25437861680984497, "learning_rate": 3.6748543081990783e-05, "loss": 0.0177, "step": 48490 }, { "grad_norm": 0.19629442691802979, "learning_rate": 3.672861493466112e-05, "loss": 0.0165, "step": 48500 }, { "grad_norm": 0.32846972346305847, "learning_rate": 3.670868905504818e-05, "loss": 0.0201, "step": 48510 }, { "grad_norm": 0.3348812758922577, "learning_rate": 3.6688765446556784e-05, "loss": 0.021, "step": 48520 }, { "grad_norm": 0.18540602922439575, "learning_rate": 3.6668844112591276e-05, "loss": 0.0211, "step": 48530 }, { "grad_norm": 0.25852155685424805, "learning_rate": 3.664892505655573e-05, "loss": 0.0156, "step": 48540 }, { "grad_norm": 0.2708718776702881, "learning_rate": 3.662900828185373e-05, "loss": 0.0217, "step": 48550 }, { "grad_norm": 0.22446046769618988, "learning_rate": 3.6609093791888516e-05, "loss": 0.0167, "step": 48560 }, { "grad_norm": 0.3401997983455658, "learning_rate": 3.658918159006292e-05, "loss": 0.0169, "step": 48570 }, { "grad_norm": 0.2954201400279999, "learning_rate": 3.656927167977942e-05, "loss": 0.0231, "step": 48580 }, { "grad_norm": 0.27529987692832947, "learning_rate": 3.654936406444006e-05, "loss": 0.0231, "step": 48590 }, { "grad_norm": 0.22336234152317047, "learning_rate": 3.65294587474465e-05, "loss": 0.013, "step": 48600 }, { "grad_norm": 0.2811508774757385, "learning_rate": 3.650955573220002e-05, "loss": 0.047, "step": 48610 }, { "grad_norm": 0.284126341342926, "learning_rate": 3.648965502210149e-05, "loss": 0.017, "step": 48620 }, { "grad_norm": 0.243904709815979, "learning_rate": 3.646975662055142e-05, "loss": 0.0176, "step": 48630 }, { "grad_norm": 0.16879385709762573, "learning_rate": 3.644986053094987e-05, "loss": 0.0172, "step": 48640 }, { "grad_norm": 0.3091984987258911, "learning_rate": 3.642996675669659e-05, "loss": 0.0253, "step": 48650 }, { "grad_norm": 0.27620929479599, "learning_rate": 3.641007530119083e-05, "loss": 0.0146, "step": 48660 }, { "grad_norm": 0.26265963912010193, "learning_rate": 3.639018616783153e-05, "loss": 0.0198, "step": 48670 }, { "grad_norm": 0.2507169842720032, "learning_rate": 3.637029936001719e-05, "loss": 0.0189, "step": 48680 }, { "grad_norm": 0.32245394587516785, "learning_rate": 3.6350414881145886e-05, "loss": 0.0268, "step": 48690 }, { "grad_norm": 0.26243749260902405, "learning_rate": 3.6330532734615386e-05, "loss": 0.0169, "step": 48700 }, { "grad_norm": 0.17870350182056427, "learning_rate": 3.6310652923822975e-05, "loss": 0.0188, "step": 48710 }, { "grad_norm": 0.21716555953025818, "learning_rate": 3.629077545216558e-05, "loss": 0.0183, "step": 48720 }, { "grad_norm": 0.30028244853019714, "learning_rate": 3.627090032303969e-05, "loss": 0.0293, "step": 48730 }, { "grad_norm": 0.2629571855068207, "learning_rate": 3.625102753984146e-05, "loss": 0.0285, "step": 48740 }, { "grad_norm": 0.2617695927619934, "learning_rate": 3.623115710596659e-05, "loss": 0.0182, "step": 48750 }, { "grad_norm": 0.42182716727256775, "learning_rate": 3.6211289024810395e-05, "loss": 0.0241, "step": 48760 }, { "grad_norm": 0.2865176796913147, "learning_rate": 3.619142329976777e-05, "loss": 0.0141, "step": 48770 }, { "grad_norm": 0.23963060975074768, "learning_rate": 3.6171559934233247e-05, "loss": 0.0158, "step": 48780 }, { "grad_norm": 0.23324726521968842, "learning_rate": 3.615169893160093e-05, "loss": 0.0182, "step": 48790 }, { "grad_norm": 0.22134386003017426, "learning_rate": 3.61318402952645e-05, "loss": 0.0198, "step": 48800 }, { "grad_norm": 0.33034300804138184, "learning_rate": 3.6111984028617285e-05, "loss": 0.0217, "step": 48810 }, { "grad_norm": 0.1757918745279312, "learning_rate": 3.6092130135052134e-05, "loss": 0.0212, "step": 48820 }, { "grad_norm": 0.2463235855102539, "learning_rate": 3.6072278617961584e-05, "loss": 0.0217, "step": 48830 }, { "grad_norm": 0.30554530024528503, "learning_rate": 3.605242948073767e-05, "loss": 0.0192, "step": 48840 }, { "grad_norm": 0.26382339000701904, "learning_rate": 3.603258272677212e-05, "loss": 0.0197, "step": 48850 }, { "grad_norm": 0.2097630351781845, "learning_rate": 3.601273835945616e-05, "loss": 0.019, "step": 48860 }, { "grad_norm": 0.2665367126464844, "learning_rate": 3.5992896382180664e-05, "loss": 0.0185, "step": 48870 }, { "grad_norm": 0.29668110609054565, "learning_rate": 3.597305679833609e-05, "loss": 0.0193, "step": 48880 }, { "grad_norm": 0.2828063666820526, "learning_rate": 3.595321961131245e-05, "loss": 0.0206, "step": 48890 }, { "grad_norm": 0.21804381906986237, "learning_rate": 3.593338482449942e-05, "loss": 0.0247, "step": 48900 }, { "grad_norm": 0.2519441843032837, "learning_rate": 3.591355244128618e-05, "loss": 0.03, "step": 48910 }, { "grad_norm": 0.30008119344711304, "learning_rate": 3.589372246506158e-05, "loss": 0.0222, "step": 48920 }, { "grad_norm": 0.2537060081958771, "learning_rate": 3.5873894899213984e-05, "loss": 0.0173, "step": 48930 }, { "grad_norm": 0.25444459915161133, "learning_rate": 3.5854069747131416e-05, "loss": 0.0163, "step": 48940 }, { "grad_norm": 0.30905836820602417, "learning_rate": 3.583424701220143e-05, "loss": 0.0183, "step": 48950 }, { "grad_norm": 0.27804258465766907, "learning_rate": 3.581442669781121e-05, "loss": 0.0221, "step": 48960 }, { "grad_norm": 0.4608222544193268, "learning_rate": 3.579460880734749e-05, "loss": 0.0208, "step": 48970 }, { "grad_norm": 0.34750354290008545, "learning_rate": 3.577479334419657e-05, "loss": 0.0248, "step": 48980 }, { "grad_norm": 0.23188534379005432, "learning_rate": 3.575498031174444e-05, "loss": 0.02, "step": 48990 }, { "grad_norm": 0.23982277512550354, "learning_rate": 3.573516971337657e-05, "loss": 0.0218, "step": 49000 }, { "grad_norm": 0.4267650842666626, "learning_rate": 3.5715361552478046e-05, "loss": 0.0206, "step": 49010 }, { "grad_norm": 0.30067920684814453, "learning_rate": 3.5695555832433536e-05, "loss": 0.0167, "step": 49020 }, { "grad_norm": 0.37195271253585815, "learning_rate": 3.5675752556627325e-05, "loss": 0.0242, "step": 49030 }, { "grad_norm": 0.3466084599494934, "learning_rate": 3.565595172844322e-05, "loss": 0.022, "step": 49040 }, { "grad_norm": 0.31790515780448914, "learning_rate": 3.5636153351264666e-05, "loss": 0.0211, "step": 49050 }, { "grad_norm": 0.2980125844478607, "learning_rate": 3.5616357428474655e-05, "loss": 0.0196, "step": 49060 }, { "grad_norm": 0.25762614607810974, "learning_rate": 3.559656396345575e-05, "loss": 0.02, "step": 49070 }, { "grad_norm": 0.20346173644065857, "learning_rate": 3.5576772959590146e-05, "loss": 0.0193, "step": 49080 }, { "grad_norm": 0.2965150475502014, "learning_rate": 3.5556984420259545e-05, "loss": 0.0182, "step": 49090 }, { "grad_norm": 0.1441548764705658, "learning_rate": 3.5537198348845305e-05, "loss": 0.0167, "step": 49100 }, { "grad_norm": 0.2237141877412796, "learning_rate": 3.551741474872831e-05, "loss": 0.0188, "step": 49110 }, { "grad_norm": 0.25827348232269287, "learning_rate": 3.549763362328903e-05, "loss": 0.0209, "step": 49120 }, { "grad_norm": 0.2970890402793884, "learning_rate": 3.5477854975907515e-05, "loss": 0.0199, "step": 49130 }, { "grad_norm": 0.21315503120422363, "learning_rate": 3.5458078809963416e-05, "loss": 0.0204, "step": 49140 }, { "grad_norm": 0.3380278944969177, "learning_rate": 3.543830512883594e-05, "loss": 0.018, "step": 49150 }, { "grad_norm": 0.2680799663066864, "learning_rate": 3.5418533935903824e-05, "loss": 0.0193, "step": 49160 }, { "grad_norm": 0.39532119035720825, "learning_rate": 3.539876523454547e-05, "loss": 0.0218, "step": 49170 }, { "grad_norm": 0.25133851170539856, "learning_rate": 3.537899902813878e-05, "loss": 0.0158, "step": 49180 }, { "grad_norm": 0.27745044231414795, "learning_rate": 3.5359235320061293e-05, "loss": 0.022, "step": 49190 }, { "grad_norm": 0.19611969590187073, "learning_rate": 3.533947411369003e-05, "loss": 0.015, "step": 49200 }, { "grad_norm": 0.17901234328746796, "learning_rate": 3.53197154124017e-05, "loss": 0.0172, "step": 49210 }, { "grad_norm": 0.19495615363121033, "learning_rate": 3.52999592195725e-05, "loss": 0.019, "step": 49220 }, { "grad_norm": 0.2897249758243561, "learning_rate": 3.5280205538578224e-05, "loss": 0.0147, "step": 49230 }, { "grad_norm": 0.3168164789676666, "learning_rate": 3.5260454372794236e-05, "loss": 0.0205, "step": 49240 }, { "grad_norm": 0.26783236861228943, "learning_rate": 3.524070572559545e-05, "loss": 0.0187, "step": 49250 }, { "grad_norm": 0.29129111766815186, "learning_rate": 3.5220959600356395e-05, "loss": 0.018, "step": 49260 }, { "grad_norm": 0.20144954323768616, "learning_rate": 3.5201216000451145e-05, "loss": 0.0153, "step": 49270 }, { "grad_norm": 0.20941030979156494, "learning_rate": 3.5181474929253335e-05, "loss": 0.0163, "step": 49280 }, { "grad_norm": 0.39923524856567383, "learning_rate": 3.516173639013615e-05, "loss": 0.018, "step": 49290 }, { "grad_norm": 0.24244646728038788, "learning_rate": 3.5142000386472406e-05, "loss": 0.0166, "step": 49300 }, { "grad_norm": 0.24345383048057556, "learning_rate": 3.5122266921634427e-05, "loss": 0.0157, "step": 49310 }, { "grad_norm": 0.29849809408187866, "learning_rate": 3.510253599899413e-05, "loss": 0.0216, "step": 49320 }, { "grad_norm": 0.20532411336898804, "learning_rate": 3.5082807621922965e-05, "loss": 0.0185, "step": 49330 }, { "grad_norm": 0.40598064661026, "learning_rate": 3.506308179379201e-05, "loss": 0.0226, "step": 49340 }, { "grad_norm": 0.2798241674900055, "learning_rate": 3.5043358517971844e-05, "loss": 0.0166, "step": 49350 }, { "grad_norm": 0.15881702303886414, "learning_rate": 3.502363779783264e-05, "loss": 0.015, "step": 49360 }, { "grad_norm": 0.33039945363998413, "learning_rate": 3.500391963674415e-05, "loss": 0.0181, "step": 49370 }, { "grad_norm": 0.4135865271091461, "learning_rate": 3.4984204038075615e-05, "loss": 0.0253, "step": 49380 }, { "grad_norm": 0.2114054262638092, "learning_rate": 3.496449100519595e-05, "loss": 0.0169, "step": 49390 }, { "grad_norm": 0.3699137270450592, "learning_rate": 3.494478054147354e-05, "loss": 0.0204, "step": 49400 }, { "grad_norm": 0.22402340173721313, "learning_rate": 3.4925072650276395e-05, "loss": 0.0161, "step": 49410 }, { "grad_norm": 0.28741347789764404, "learning_rate": 3.4905367334972016e-05, "loss": 0.0154, "step": 49420 }, { "grad_norm": 0.41568344831466675, "learning_rate": 3.488566459892752e-05, "loss": 0.0241, "step": 49430 }, { "grad_norm": 0.3353424370288849, "learning_rate": 3.4865964445509585e-05, "loss": 0.0202, "step": 49440 }, { "grad_norm": 0.2901501953601837, "learning_rate": 3.484626687808438e-05, "loss": 0.0183, "step": 49450 }, { "grad_norm": 0.2599909007549286, "learning_rate": 3.4826571900017735e-05, "loss": 0.0142, "step": 49460 }, { "grad_norm": 0.5691324472427368, "learning_rate": 3.480687951467495e-05, "loss": 0.0153, "step": 49470 }, { "grad_norm": 0.22873902320861816, "learning_rate": 3.4787189725420925e-05, "loss": 0.0201, "step": 49480 }, { "grad_norm": 0.19530938565731049, "learning_rate": 3.4767502535620086e-05, "loss": 0.0197, "step": 49490 }, { "grad_norm": 0.3306911289691925, "learning_rate": 3.474781794863648e-05, "loss": 0.0248, "step": 49500 }, { "grad_norm": 0.44543570280075073, "learning_rate": 3.472813596783363e-05, "loss": 0.0221, "step": 49510 }, { "grad_norm": 0.2761762738227844, "learning_rate": 3.470845659657466e-05, "loss": 0.021, "step": 49520 }, { "grad_norm": 0.30783137679100037, "learning_rate": 3.468877983822223e-05, "loss": 0.0165, "step": 49530 }, { "grad_norm": 0.37914445996284485, "learning_rate": 3.466910569613855e-05, "loss": 0.02, "step": 49540 }, { "grad_norm": 0.27764469385147095, "learning_rate": 3.464943417368542e-05, "loss": 0.0245, "step": 49550 }, { "grad_norm": 0.2717466354370117, "learning_rate": 3.462976527422415e-05, "loss": 0.0176, "step": 49560 }, { "grad_norm": 0.38896462321281433, "learning_rate": 3.461009900111562e-05, "loss": 0.0239, "step": 49570 }, { "grad_norm": 0.4530760943889618, "learning_rate": 3.459043535772023e-05, "loss": 0.0218, "step": 49580 }, { "grad_norm": 0.33008942008018494, "learning_rate": 3.4570774347398014e-05, "loss": 0.0237, "step": 49590 }, { "grad_norm": 0.33219531178474426, "learning_rate": 3.4551115973508454e-05, "loss": 0.0211, "step": 49600 }, { "grad_norm": 0.3294844627380371, "learning_rate": 3.453146023941066e-05, "loss": 0.0225, "step": 49610 }, { "grad_norm": 0.33817610144615173, "learning_rate": 3.451180714846325e-05, "loss": 0.0222, "step": 49620 }, { "grad_norm": 0.4137658476829529, "learning_rate": 3.449215670402438e-05, "loss": 0.0188, "step": 49630 }, { "grad_norm": 0.3849474787712097, "learning_rate": 3.447250890945181e-05, "loss": 0.0188, "step": 49640 }, { "grad_norm": 0.2769134044647217, "learning_rate": 3.4452863768102754e-05, "loss": 0.0174, "step": 49650 }, { "grad_norm": 0.2865539491176605, "learning_rate": 3.443322128333409e-05, "loss": 0.0218, "step": 49660 }, { "grad_norm": 0.3860444724559784, "learning_rate": 3.441358145850215e-05, "loss": 0.0206, "step": 49670 }, { "grad_norm": 0.2319997102022171, "learning_rate": 3.439394429696286e-05, "loss": 0.0226, "step": 49680 }, { "grad_norm": 0.3038254976272583, "learning_rate": 3.4374309802071644e-05, "loss": 0.0177, "step": 49690 }, { "grad_norm": 0.22061781585216522, "learning_rate": 3.435467797718353e-05, "loss": 0.0172, "step": 49700 }, { "grad_norm": 0.20405326783657074, "learning_rate": 3.433504882565306e-05, "loss": 0.0203, "step": 49710 }, { "grad_norm": 0.19993530213832855, "learning_rate": 3.43154223508343e-05, "loss": 0.0213, "step": 49720 }, { "grad_norm": 0.3171815574169159, "learning_rate": 3.429579855608089e-05, "loss": 0.0173, "step": 49730 }, { "grad_norm": 0.27344605326652527, "learning_rate": 3.427617744474597e-05, "loss": 0.0187, "step": 49740 }, { "grad_norm": 0.32919803261756897, "learning_rate": 3.425655902018231e-05, "loss": 0.0221, "step": 49750 }, { "grad_norm": 0.27925726771354675, "learning_rate": 3.423694328574211e-05, "loss": 0.0161, "step": 49760 }, { "grad_norm": 0.37921786308288574, "learning_rate": 3.42173302447772e-05, "loss": 0.0214, "step": 49770 }, { "grad_norm": 0.22479553520679474, "learning_rate": 3.419771990063886e-05, "loss": 0.015, "step": 49780 }, { "grad_norm": 0.15383654832839966, "learning_rate": 3.417811225667803e-05, "loss": 0.0144, "step": 49790 }, { "grad_norm": 0.2964613139629364, "learning_rate": 3.415850731624508e-05, "loss": 0.0204, "step": 49800 }, { "grad_norm": 0.37436044216156006, "learning_rate": 3.4138905082689945e-05, "loss": 0.0224, "step": 49810 }, { "grad_norm": 0.2545638084411621, "learning_rate": 3.4119305559362145e-05, "loss": 0.0223, "step": 49820 }, { "grad_norm": 0.3323783278465271, "learning_rate": 3.4099708749610684e-05, "loss": 0.0206, "step": 49830 }, { "grad_norm": 0.3017547130584717, "learning_rate": 3.408011465678413e-05, "loss": 0.0257, "step": 49840 }, { "grad_norm": 0.2694622874259949, "learning_rate": 3.406052328423055e-05, "loss": 0.0224, "step": 49850 }, { "grad_norm": 0.6568906903266907, "learning_rate": 3.4040934635297615e-05, "loss": 0.0197, "step": 49860 }, { "grad_norm": 0.30039501190185547, "learning_rate": 3.4021348713332466e-05, "loss": 0.0162, "step": 49870 }, { "grad_norm": 0.26357731223106384, "learning_rate": 3.4001765521681807e-05, "loss": 0.0179, "step": 49880 }, { "grad_norm": 0.24994981288909912, "learning_rate": 3.398218506369188e-05, "loss": 0.0262, "step": 49890 }, { "grad_norm": 0.2896038591861725, "learning_rate": 3.3962607342708404e-05, "loss": 0.0203, "step": 49900 }, { "grad_norm": 0.2593287527561188, "learning_rate": 3.394303236207673e-05, "loss": 0.0184, "step": 49910 }, { "grad_norm": 0.23276247084140778, "learning_rate": 3.392346012514166e-05, "loss": 0.0188, "step": 49920 }, { "grad_norm": 0.25041258335113525, "learning_rate": 3.390389063524757e-05, "loss": 0.014, "step": 49930 }, { "grad_norm": 0.20305147767066956, "learning_rate": 3.3884323895738324e-05, "loss": 0.015, "step": 49940 }, { "grad_norm": 0.36599287390708923, "learning_rate": 3.386475990995738e-05, "loss": 0.0235, "step": 49950 }, { "grad_norm": 0.4697551429271698, "learning_rate": 3.384519868124765e-05, "loss": 0.0178, "step": 49960 }, { "grad_norm": 0.2307426929473877, "learning_rate": 3.3825640212951645e-05, "loss": 0.0171, "step": 49970 }, { "grad_norm": 0.22468218207359314, "learning_rate": 3.380608450841134e-05, "loss": 0.0176, "step": 49980 }, { "grad_norm": 0.3458351194858551, "learning_rate": 3.3786531570968305e-05, "loss": 0.0187, "step": 49990 }, { "grad_norm": 0.20777954161167145, "learning_rate": 3.3766981403963584e-05, "loss": 0.0189, "step": 50000 }, { "grad_norm": 0.40947240591049194, "learning_rate": 3.374743401073775e-05, "loss": 0.0191, "step": 50010 }, { "grad_norm": 0.36319899559020996, "learning_rate": 3.372788939463095e-05, "loss": 0.0178, "step": 50020 }, { "grad_norm": 0.2997684180736542, "learning_rate": 3.370834755898281e-05, "loss": 0.0169, "step": 50030 }, { "grad_norm": 0.4183867275714874, "learning_rate": 3.3688808507132493e-05, "loss": 0.0204, "step": 50040 }, { "grad_norm": 0.23538771271705627, "learning_rate": 3.3669272242418685e-05, "loss": 0.0161, "step": 50050 }, { "grad_norm": 0.21142564713954926, "learning_rate": 3.364973876817961e-05, "loss": 0.0186, "step": 50060 }, { "grad_norm": 0.5351308584213257, "learning_rate": 3.363020808775299e-05, "loss": 0.02, "step": 50070 }, { "grad_norm": 0.24117696285247803, "learning_rate": 3.361068020447611e-05, "loss": 0.0134, "step": 50080 }, { "grad_norm": 0.3003462553024292, "learning_rate": 3.3591155121685724e-05, "loss": 0.0166, "step": 50090 }, { "grad_norm": 0.2679234743118286, "learning_rate": 3.357163284271814e-05, "loss": 0.0189, "step": 50100 }, { "grad_norm": 0.35499632358551025, "learning_rate": 3.355211337090919e-05, "loss": 0.0187, "step": 50110 }, { "grad_norm": 0.30994871258735657, "learning_rate": 3.353259670959421e-05, "loss": 0.0221, "step": 50120 }, { "grad_norm": 0.34060966968536377, "learning_rate": 3.351308286210808e-05, "loss": 0.0196, "step": 50130 }, { "grad_norm": 0.2335607260465622, "learning_rate": 3.3493571831785156e-05, "loss": 0.0155, "step": 50140 }, { "grad_norm": 0.2916705310344696, "learning_rate": 3.347406362195936e-05, "loss": 0.0182, "step": 50150 }, { "grad_norm": 0.23883812129497528, "learning_rate": 3.345455823596411e-05, "loss": 0.0218, "step": 50160 }, { "grad_norm": 0.43102240562438965, "learning_rate": 3.3435055677132346e-05, "loss": 0.0184, "step": 50170 }, { "grad_norm": 0.3433171510696411, "learning_rate": 3.3415555948796505e-05, "loss": 0.018, "step": 50180 }, { "grad_norm": 0.3216637969017029, "learning_rate": 3.3396059054288556e-05, "loss": 0.017, "step": 50190 }, { "grad_norm": 0.20011647045612335, "learning_rate": 3.3376564996940015e-05, "loss": 0.0276, "step": 50200 }, { "grad_norm": 0.21201476454734802, "learning_rate": 3.3357073780081836e-05, "loss": 0.0155, "step": 50210 }, { "grad_norm": 0.2779257893562317, "learning_rate": 3.333758540704459e-05, "loss": 0.017, "step": 50220 }, { "grad_norm": 0.3390922248363495, "learning_rate": 3.331809988115827e-05, "loss": 0.0186, "step": 50230 }, { "grad_norm": 0.2848704159259796, "learning_rate": 3.329861720575244e-05, "loss": 0.0183, "step": 50240 }, { "grad_norm": 0.30082058906555176, "learning_rate": 3.3279137384156126e-05, "loss": 0.0132, "step": 50250 }, { "grad_norm": 0.3961150348186493, "learning_rate": 3.3259660419697934e-05, "loss": 0.0211, "step": 50260 }, { "grad_norm": 0.29361456632614136, "learning_rate": 3.3240186315705926e-05, "loss": 0.0193, "step": 50270 }, { "grad_norm": 0.16953983902931213, "learning_rate": 3.322071507550769e-05, "loss": 0.0197, "step": 50280 }, { "grad_norm": 0.23334386944770813, "learning_rate": 3.320124670243033e-05, "loss": 0.0188, "step": 50290 }, { "grad_norm": 0.19966745376586914, "learning_rate": 3.318178119980045e-05, "loss": 0.0153, "step": 50300 }, { "grad_norm": 0.26659414172172546, "learning_rate": 3.316231857094421e-05, "loss": 0.0212, "step": 50310 }, { "grad_norm": 0.30549630522727966, "learning_rate": 3.31428588191872e-05, "loss": 0.0225, "step": 50320 }, { "grad_norm": 0.4356749653816223, "learning_rate": 3.312340194785458e-05, "loss": 0.0203, "step": 50330 }, { "grad_norm": 0.3096257150173187, "learning_rate": 3.3103947960270975e-05, "loss": 0.0183, "step": 50340 }, { "grad_norm": 0.36284980177879333, "learning_rate": 3.308449685976058e-05, "loss": 0.0158, "step": 50350 }, { "grad_norm": 0.2270132601261139, "learning_rate": 3.3065048649647024e-05, "loss": 0.0168, "step": 50360 }, { "grad_norm": 0.5952246785163879, "learning_rate": 3.304560333325348e-05, "loss": 0.0212, "step": 50370 }, { "grad_norm": 0.36101916432380676, "learning_rate": 3.3026160913902624e-05, "loss": 0.0189, "step": 50380 }, { "grad_norm": 0.4128076732158661, "learning_rate": 3.300672139491662e-05, "loss": 0.0261, "step": 50390 }, { "grad_norm": 0.22752611339092255, "learning_rate": 3.298728477961717e-05, "loss": 0.0163, "step": 50400 }, { "grad_norm": 0.21370524168014526, "learning_rate": 3.296785107132544e-05, "loss": 0.0247, "step": 50410 }, { "grad_norm": 0.3568902611732483, "learning_rate": 3.2948420273362156e-05, "loss": 0.0188, "step": 50420 }, { "grad_norm": 0.29243743419647217, "learning_rate": 3.292899238904747e-05, "loss": 0.0183, "step": 50430 }, { "grad_norm": 0.23574694991111755, "learning_rate": 3.29095674217011e-05, "loss": 0.0193, "step": 50440 }, { "grad_norm": 0.22327564656734467, "learning_rate": 3.289014537464224e-05, "loss": 0.0183, "step": 50450 }, { "grad_norm": 0.26453647017478943, "learning_rate": 3.287072625118955e-05, "loss": 0.0181, "step": 50460 }, { "grad_norm": 0.3363019824028015, "learning_rate": 3.285131005466129e-05, "loss": 0.0148, "step": 50470 }, { "grad_norm": 0.24374568462371826, "learning_rate": 3.2831896788375105e-05, "loss": 0.0153, "step": 50480 }, { "grad_norm": 0.2879171371459961, "learning_rate": 3.281248645564822e-05, "loss": 0.0188, "step": 50490 }, { "grad_norm": 0.1394195407629013, "learning_rate": 3.2793079059797306e-05, "loss": 0.0229, "step": 50500 }, { "grad_norm": 0.20716755092144012, "learning_rate": 3.277367460413859e-05, "loss": 0.017, "step": 50510 }, { "grad_norm": 0.36482134461402893, "learning_rate": 3.275427309198773e-05, "loss": 0.0237, "step": 50520 }, { "grad_norm": 0.201946422457695, "learning_rate": 3.273487452665993e-05, "loss": 0.0177, "step": 50530 }, { "grad_norm": 0.3026796579360962, "learning_rate": 3.271547891146986e-05, "loss": 0.0186, "step": 50540 }, { "grad_norm": 0.2998964786529541, "learning_rate": 3.269608624973173e-05, "loss": 0.0166, "step": 50550 }, { "grad_norm": 0.39251455664634705, "learning_rate": 3.26766965447592e-05, "loss": 0.0202, "step": 50560 }, { "grad_norm": 0.27294856309890747, "learning_rate": 3.2657309799865424e-05, "loss": 0.0182, "step": 50570 }, { "grad_norm": 1.1908714771270752, "learning_rate": 3.2637926018363084e-05, "loss": 0.0215, "step": 50580 }, { "grad_norm": 0.3055752217769623, "learning_rate": 3.261854520356432e-05, "loss": 0.018, "step": 50590 }, { "grad_norm": 0.13987962901592255, "learning_rate": 3.2599167358780816e-05, "loss": 0.0164, "step": 50600 }, { "grad_norm": 0.4304838180541992, "learning_rate": 3.2579792487323676e-05, "loss": 0.0137, "step": 50610 }, { "grad_norm": 0.3129092752933502, "learning_rate": 3.256042059250358e-05, "loss": 0.0191, "step": 50620 }, { "grad_norm": 0.5214062333106995, "learning_rate": 3.254105167763062e-05, "loss": 0.0285, "step": 50630 }, { "grad_norm": 0.2547711730003357, "learning_rate": 3.252168574601443e-05, "loss": 0.0255, "step": 50640 }, { "grad_norm": 0.17092178761959076, "learning_rate": 3.2502322800964115e-05, "loss": 0.0172, "step": 50650 }, { "grad_norm": 0.3385734558105469, "learning_rate": 3.248296284578825e-05, "loss": 0.016, "step": 50660 }, { "grad_norm": 0.25301289558410645, "learning_rate": 3.246360588379497e-05, "loss": 0.018, "step": 50670 }, { "grad_norm": 0.3829041123390198, "learning_rate": 3.24442519182918e-05, "loss": 0.0343, "step": 50680 }, { "grad_norm": 0.1908254623413086, "learning_rate": 3.2424900952585845e-05, "loss": 0.0142, "step": 50690 }, { "grad_norm": 0.36010196805000305, "learning_rate": 3.240555298998362e-05, "loss": 0.0176, "step": 50700 }, { "grad_norm": 0.2919692397117615, "learning_rate": 3.2386208033791194e-05, "loss": 0.0177, "step": 50710 }, { "grad_norm": 0.24187366664409637, "learning_rate": 3.2366866087314064e-05, "loss": 0.0182, "step": 50720 }, { "grad_norm": 0.3272153437137604, "learning_rate": 3.234752715385727e-05, "loss": 0.0213, "step": 50730 }, { "grad_norm": 0.2392553985118866, "learning_rate": 3.232819123672529e-05, "loss": 0.0238, "step": 50740 }, { "grad_norm": 0.41170722246170044, "learning_rate": 3.2308858339222076e-05, "loss": 0.0223, "step": 50750 }, { "grad_norm": 0.17949528992176056, "learning_rate": 3.2289528464651144e-05, "loss": 0.0177, "step": 50760 }, { "grad_norm": 0.18588954210281372, "learning_rate": 3.22702016163154e-05, "loss": 0.0209, "step": 50770 }, { "grad_norm": 0.23099879920482635, "learning_rate": 3.2250877797517295e-05, "loss": 0.0174, "step": 50780 }, { "grad_norm": 0.2901146411895752, "learning_rate": 3.223155701155872e-05, "loss": 0.0204, "step": 50790 }, { "grad_norm": 0.29295873641967773, "learning_rate": 3.2212239261741105e-05, "loss": 0.0277, "step": 50800 }, { "grad_norm": 0.38221603631973267, "learning_rate": 3.219292455136528e-05, "loss": 0.0168, "step": 50810 }, { "grad_norm": 0.20921534299850464, "learning_rate": 3.217361288373165e-05, "loss": 0.016, "step": 50820 }, { "grad_norm": 0.11881416290998459, "learning_rate": 3.215430426214002e-05, "loss": 0.0211, "step": 50830 }, { "grad_norm": 0.19814173877239227, "learning_rate": 3.2134998689889705e-05, "loss": 0.0136, "step": 50840 }, { "grad_norm": 0.20099897682666779, "learning_rate": 3.211569617027952e-05, "loss": 0.0195, "step": 50850 }, { "grad_norm": 0.2557518184185028, "learning_rate": 3.209639670660769e-05, "loss": 0.0189, "step": 50860 }, { "grad_norm": 0.48125210404396057, "learning_rate": 3.207710030217202e-05, "loss": 0.0187, "step": 50870 }, { "grad_norm": 0.2674781084060669, "learning_rate": 3.205780696026972e-05, "loss": 0.016, "step": 50880 }, { "grad_norm": 0.43193504214286804, "learning_rate": 3.203851668419749e-05, "loss": 0.016, "step": 50890 }, { "grad_norm": 0.2986830174922943, "learning_rate": 3.201922947725149e-05, "loss": 0.0169, "step": 50900 }, { "grad_norm": 0.5008911490440369, "learning_rate": 3.199994534272742e-05, "loss": 0.0163, "step": 50910 }, { "grad_norm": 0.20709389448165894, "learning_rate": 3.1980664283920394e-05, "loss": 0.0223, "step": 50920 }, { "grad_norm": 0.2804844379425049, "learning_rate": 3.196138630412499e-05, "loss": 0.0177, "step": 50930 }, { "grad_norm": 0.5380629897117615, "learning_rate": 3.194211140663532e-05, "loss": 0.0169, "step": 50940 }, { "grad_norm": 0.32942643761634827, "learning_rate": 3.1922839594744914e-05, "loss": 0.0218, "step": 50950 }, { "grad_norm": 0.2065792828798294, "learning_rate": 3.190357087174683e-05, "loss": 0.0214, "step": 50960 }, { "grad_norm": 0.1933521330356598, "learning_rate": 3.188430524093353e-05, "loss": 0.0153, "step": 50970 }, { "grad_norm": 0.26739925146102905, "learning_rate": 3.1865042705597014e-05, "loss": 0.0209, "step": 50980 }, { "grad_norm": 0.2638365626335144, "learning_rate": 3.184578326902868e-05, "loss": 0.0203, "step": 50990 }, { "grad_norm": 0.22741976380348206, "learning_rate": 3.182652693451949e-05, "loss": 0.0201, "step": 51000 }, { "grad_norm": 0.36011090874671936, "learning_rate": 3.180727370535981e-05, "loss": 0.0171, "step": 51010 }, { "grad_norm": 0.23983319103717804, "learning_rate": 3.178802358483945e-05, "loss": 0.0177, "step": 51020 }, { "grad_norm": 0.2798464298248291, "learning_rate": 3.17687765762478e-05, "loss": 0.0149, "step": 51030 }, { "grad_norm": 0.30664288997650146, "learning_rate": 3.174953268287357e-05, "loss": 0.0213, "step": 51040 }, { "grad_norm": 0.33581840991973877, "learning_rate": 3.173029190800508e-05, "loss": 0.0144, "step": 51050 }, { "grad_norm": 0.259246289730072, "learning_rate": 3.171105425493e-05, "loss": 0.0185, "step": 51060 }, { "grad_norm": 0.19531334936618805, "learning_rate": 3.169181972693557e-05, "loss": 0.0136, "step": 51070 }, { "grad_norm": 0.22725068032741547, "learning_rate": 3.1672588327308404e-05, "loss": 0.0141, "step": 51080 }, { "grad_norm": 0.40598171949386597, "learning_rate": 3.1653360059334646e-05, "loss": 0.0186, "step": 51090 }, { "grad_norm": 0.510657012462616, "learning_rate": 3.163413492629985e-05, "loss": 0.0278, "step": 51100 }, { "grad_norm": 0.30427637696266174, "learning_rate": 3.161491293148912e-05, "loss": 0.0192, "step": 51110 }, { "grad_norm": 0.21394050121307373, "learning_rate": 3.1595694078186925e-05, "loss": 0.0185, "step": 51120 }, { "grad_norm": 0.40023571252822876, "learning_rate": 3.157647836967724e-05, "loss": 0.0166, "step": 51130 }, { "grad_norm": 0.37129855155944824, "learning_rate": 3.155726580924354e-05, "loss": 0.0174, "step": 51140 }, { "grad_norm": 0.32271280884742737, "learning_rate": 3.1538056400168676e-05, "loss": 0.0279, "step": 51150 }, { "grad_norm": 0.2188028246164322, "learning_rate": 3.151885014573506e-05, "loss": 0.0285, "step": 51160 }, { "grad_norm": 0.3357948958873749, "learning_rate": 3.149964704922448e-05, "loss": 0.028, "step": 51170 }, { "grad_norm": 0.2666487693786621, "learning_rate": 3.148044711391825e-05, "loss": 0.0161, "step": 51180 }, { "grad_norm": 0.3077227473258972, "learning_rate": 3.146125034309707e-05, "loss": 0.0201, "step": 51190 }, { "grad_norm": 0.1835608184337616, "learning_rate": 3.1442056740041195e-05, "loss": 0.024, "step": 51200 }, { "grad_norm": 0.280453085899353, "learning_rate": 3.142286630803026e-05, "loss": 0.0315, "step": 51210 }, { "grad_norm": 0.19684286415576935, "learning_rate": 3.140367905034337e-05, "loss": 0.019, "step": 51220 }, { "grad_norm": 0.22572200000286102, "learning_rate": 3.138449497025914e-05, "loss": 0.0176, "step": 51230 }, { "grad_norm": 0.3618084192276001, "learning_rate": 3.136531407105557e-05, "loss": 0.0176, "step": 51240 }, { "grad_norm": 0.21833515167236328, "learning_rate": 3.1346136356010184e-05, "loss": 0.0168, "step": 51250 }, { "grad_norm": 0.1475014090538025, "learning_rate": 3.132696182839988e-05, "loss": 0.0159, "step": 51260 }, { "grad_norm": 0.2716582715511322, "learning_rate": 3.130779049150111e-05, "loss": 0.0203, "step": 51270 }, { "grad_norm": 0.25756973028182983, "learning_rate": 3.128862234858971e-05, "loss": 0.0199, "step": 51280 }, { "grad_norm": 0.18253234028816223, "learning_rate": 3.1269457402941005e-05, "loss": 0.0179, "step": 51290 }, { "grad_norm": 0.1741362065076828, "learning_rate": 3.125029565782974e-05, "loss": 0.0181, "step": 51300 }, { "grad_norm": 0.25477832555770874, "learning_rate": 3.123113711653013e-05, "loss": 0.0164, "step": 51310 }, { "grad_norm": 0.21660983562469482, "learning_rate": 3.121198178231587e-05, "loss": 0.0204, "step": 51320 }, { "grad_norm": 0.34506434202194214, "learning_rate": 3.119282965846006e-05, "loss": 0.0189, "step": 51330 }, { "grad_norm": 0.30515027046203613, "learning_rate": 3.1173680748235303e-05, "loss": 0.0235, "step": 51340 }, { "grad_norm": 0.23921525478363037, "learning_rate": 3.115453505491358e-05, "loss": 0.0201, "step": 51350 }, { "grad_norm": 0.3526884615421295, "learning_rate": 3.1135392581766404e-05, "loss": 0.0211, "step": 51360 }, { "grad_norm": 0.3569287061691284, "learning_rate": 3.111625333206467e-05, "loss": 0.0152, "step": 51370 }, { "grad_norm": 0.15404683351516724, "learning_rate": 3.109711730907881e-05, "loss": 0.0134, "step": 51380 }, { "grad_norm": 0.28017738461494446, "learning_rate": 3.107798451607859e-05, "loss": 0.0163, "step": 51390 }, { "grad_norm": 0.29372116923332214, "learning_rate": 3.105885495633328e-05, "loss": 0.0202, "step": 51400 }, { "grad_norm": 0.13371242582798004, "learning_rate": 3.103972863311163e-05, "loss": 0.0142, "step": 51410 }, { "grad_norm": 0.24696888029575348, "learning_rate": 3.102060554968178e-05, "loss": 0.0173, "step": 51420 }, { "grad_norm": 0.343199223279953, "learning_rate": 3.100148570931137e-05, "loss": 0.0191, "step": 51430 }, { "grad_norm": 0.3176901638507843, "learning_rate": 3.098236911526744e-05, "loss": 0.0175, "step": 51440 }, { "grad_norm": 0.2731756865978241, "learning_rate": 3.0963255770816495e-05, "loss": 0.0161, "step": 51450 }, { "grad_norm": 0.2723148763179779, "learning_rate": 3.094414567922447e-05, "loss": 0.0191, "step": 51460 }, { "grad_norm": 0.18395549058914185, "learning_rate": 3.092503884375678e-05, "loss": 0.0189, "step": 51470 }, { "grad_norm": 0.5290138721466064, "learning_rate": 3.0905935267678255e-05, "loss": 0.0185, "step": 51480 }, { "grad_norm": 0.1580248773097992, "learning_rate": 3.0886834954253166e-05, "loss": 0.0206, "step": 51490 }, { "grad_norm": 0.30435308814048767, "learning_rate": 3.086773790674524e-05, "loss": 0.0195, "step": 51500 }, { "grad_norm": 0.23008686304092407, "learning_rate": 3.0848644128417617e-05, "loss": 0.0178, "step": 51510 }, { "grad_norm": 0.21653032302856445, "learning_rate": 3.082955362253294e-05, "loss": 0.023, "step": 51520 }, { "grad_norm": 0.36306601762771606, "learning_rate": 3.081046639235323e-05, "loss": 0.0192, "step": 51530 }, { "grad_norm": 0.257644385099411, "learning_rate": 3.0791382441139986e-05, "loss": 0.0142, "step": 51540 }, { "grad_norm": 0.2366047501564026, "learning_rate": 3.077230177215411e-05, "loss": 0.0244, "step": 51550 }, { "grad_norm": 0.1687120944261551, "learning_rate": 3.0753224388656e-05, "loss": 0.014, "step": 51560 }, { "grad_norm": 0.14834171533584595, "learning_rate": 3.073415029390544e-05, "loss": 0.0164, "step": 51570 }, { "grad_norm": 0.23071333765983582, "learning_rate": 3.071507949116166e-05, "loss": 0.016, "step": 51580 }, { "grad_norm": 0.3211451470851898, "learning_rate": 3.069601198368337e-05, "loss": 0.0141, "step": 51590 }, { "grad_norm": 0.23333413898944855, "learning_rate": 3.067694777472864e-05, "loss": 0.0133, "step": 51600 }, { "grad_norm": 0.3846494257450104, "learning_rate": 3.065788686755508e-05, "loss": 0.0197, "step": 51610 }, { "grad_norm": 0.2596226632595062, "learning_rate": 3.063882926541961e-05, "loss": 0.0157, "step": 51620 }, { "grad_norm": 0.3331490159034729, "learning_rate": 3.061977497157872e-05, "loss": 0.0173, "step": 51630 }, { "grad_norm": 0.30533909797668457, "learning_rate": 3.0600723989288224e-05, "loss": 0.0178, "step": 51640 }, { "grad_norm": 0.2998269498348236, "learning_rate": 3.058167632180343e-05, "loss": 0.0181, "step": 51650 }, { "grad_norm": 0.20554959774017334, "learning_rate": 3.0562631972379074e-05, "loss": 0.0177, "step": 51660 }, { "grad_norm": 0.2931155264377594, "learning_rate": 3.0543590944269276e-05, "loss": 0.0177, "step": 51670 }, { "grad_norm": 0.5187722444534302, "learning_rate": 3.052455324072766e-05, "loss": 0.0267, "step": 51680 }, { "grad_norm": 0.34254220128059387, "learning_rate": 3.0505518865007244e-05, "loss": 0.0158, "step": 51690 }, { "grad_norm": 0.2294444888830185, "learning_rate": 3.048648782036048e-05, "loss": 0.0156, "step": 51700 }, { "grad_norm": 0.1753394603729248, "learning_rate": 3.0467460110039236e-05, "loss": 0.0199, "step": 51710 }, { "grad_norm": 0.23079928755760193, "learning_rate": 3.0448435737294866e-05, "loss": 0.0179, "step": 51720 }, { "grad_norm": 0.20306147634983063, "learning_rate": 3.042941470537808e-05, "loss": 0.0151, "step": 51730 }, { "grad_norm": 0.266915500164032, "learning_rate": 3.0410397017539072e-05, "loss": 0.0159, "step": 51740 }, { "grad_norm": 0.2701738476753235, "learning_rate": 3.039138267702742e-05, "loss": 0.0156, "step": 51750 }, { "grad_norm": 0.21899153292179108, "learning_rate": 3.0372371687092193e-05, "loss": 0.0169, "step": 51760 }, { "grad_norm": 0.333783894777298, "learning_rate": 3.0353364050981823e-05, "loss": 0.0208, "step": 51770 }, { "grad_norm": 0.42700663208961487, "learning_rate": 3.033435977194418e-05, "loss": 0.0175, "step": 51780 }, { "grad_norm": 0.16393691301345825, "learning_rate": 3.0315358853226618e-05, "loss": 0.0143, "step": 51790 }, { "grad_norm": 0.14956247806549072, "learning_rate": 3.0296361298075844e-05, "loss": 0.0176, "step": 51800 }, { "grad_norm": 0.15532180666923523, "learning_rate": 3.027736710973803e-05, "loss": 0.0193, "step": 51810 }, { "grad_norm": 0.20642244815826416, "learning_rate": 3.025837629145875e-05, "loss": 0.0136, "step": 51820 }, { "grad_norm": 0.21403160691261292, "learning_rate": 3.0239388846483048e-05, "loss": 0.019, "step": 51830 }, { "grad_norm": 0.270485520362854, "learning_rate": 3.022040477805532e-05, "loss": 0.0225, "step": 51840 }, { "grad_norm": 0.34734073281288147, "learning_rate": 3.020142408941946e-05, "loss": 0.015, "step": 51850 }, { "grad_norm": 0.21838918328285217, "learning_rate": 3.018244678381873e-05, "loss": 0.0173, "step": 51860 }, { "grad_norm": 0.42837730050086975, "learning_rate": 3.0163472864495812e-05, "loss": 0.0229, "step": 51870 }, { "grad_norm": 0.25032052397727966, "learning_rate": 3.014450233469287e-05, "loss": 0.0175, "step": 51880 }, { "grad_norm": 0.24646802246570587, "learning_rate": 3.0125535197651422e-05, "loss": 0.0212, "step": 51890 }, { "grad_norm": 0.19203385710716248, "learning_rate": 3.0106571456612448e-05, "loss": 0.0133, "step": 51900 }, { "grad_norm": 0.34681078791618347, "learning_rate": 3.0087611114816305e-05, "loss": 0.0239, "step": 51910 }, { "grad_norm": 0.16472230851650238, "learning_rate": 3.006865417550284e-05, "loss": 0.0146, "step": 51920 }, { "grad_norm": 0.26822224259376526, "learning_rate": 3.0049700641911242e-05, "loss": 0.0161, "step": 51930 }, { "grad_norm": 0.16210469603538513, "learning_rate": 3.0030750517280175e-05, "loss": 0.0201, "step": 51940 }, { "grad_norm": 0.1820049285888672, "learning_rate": 3.0011803804847682e-05, "loss": 0.02, "step": 51950 }, { "grad_norm": 0.5048683881759644, "learning_rate": 2.9992860507851224e-05, "loss": 0.019, "step": 51960 }, { "grad_norm": 0.2843330502510071, "learning_rate": 2.9973920629527723e-05, "loss": 0.0166, "step": 51970 }, { "grad_norm": 0.1591774821281433, "learning_rate": 2.9954984173113453e-05, "loss": 0.0184, "step": 51980 }, { "grad_norm": 0.7333458662033081, "learning_rate": 2.993605114184418e-05, "loss": 0.0223, "step": 51990 }, { "grad_norm": 0.3713201582431793, "learning_rate": 2.9917121538955005e-05, "loss": 0.0249, "step": 52000 }, { "grad_norm": 0.36315011978149414, "learning_rate": 2.98981953676805e-05, "loss": 0.0271, "step": 52010 }, { "grad_norm": 0.2089996635913849, "learning_rate": 2.9879272631254594e-05, "loss": 0.0176, "step": 52020 }, { "grad_norm": 0.23603354394435883, "learning_rate": 2.986035333291073e-05, "loss": 0.0216, "step": 52030 }, { "grad_norm": 0.1584487110376358, "learning_rate": 2.9841437475881646e-05, "loss": 0.0221, "step": 52040 }, { "grad_norm": 0.23508144915103912, "learning_rate": 2.9822525063399555e-05, "loss": 0.0172, "step": 52050 }, { "grad_norm": 0.5081542730331421, "learning_rate": 2.9803616098696087e-05, "loss": 0.0179, "step": 52060 }, { "grad_norm": 0.2432495802640915, "learning_rate": 2.9784710585002228e-05, "loss": 0.0154, "step": 52070 }, { "grad_norm": 0.501215934753418, "learning_rate": 2.9765808525548467e-05, "loss": 0.0152, "step": 52080 }, { "grad_norm": 0.2352469265460968, "learning_rate": 2.974690992356461e-05, "loss": 0.0165, "step": 52090 }, { "grad_norm": 0.24351118505001068, "learning_rate": 2.9728014782279934e-05, "loss": 0.0216, "step": 52100 }, { "grad_norm": 0.46165648102760315, "learning_rate": 2.970912310492307e-05, "loss": 0.0166, "step": 52110 }, { "grad_norm": 0.21198421716690063, "learning_rate": 2.9690234894722134e-05, "loss": 0.0204, "step": 52120 }, { "grad_norm": 0.2523851692676544, "learning_rate": 2.9671350154904577e-05, "loss": 0.0188, "step": 52130 }, { "grad_norm": 0.25637784600257874, "learning_rate": 2.9652468888697282e-05, "loss": 0.0177, "step": 52140 }, { "grad_norm": 0.27519845962524414, "learning_rate": 2.9633591099326562e-05, "loss": 0.016, "step": 52150 }, { "grad_norm": 0.35261428356170654, "learning_rate": 2.9614716790018072e-05, "loss": 0.0172, "step": 52160 }, { "grad_norm": 0.3415299654006958, "learning_rate": 2.959584596399697e-05, "loss": 0.0174, "step": 52170 }, { "grad_norm": 0.3276090621948242, "learning_rate": 2.9576978624487717e-05, "loss": 0.0169, "step": 52180 }, { "grad_norm": 0.28454697132110596, "learning_rate": 2.9558114774714263e-05, "loss": 0.0188, "step": 52190 }, { "grad_norm": 0.28624868392944336, "learning_rate": 2.9539254417899897e-05, "loss": 0.0203, "step": 52200 }, { "grad_norm": 0.2574732005596161, "learning_rate": 2.9520397557267365e-05, "loss": 0.0128, "step": 52210 }, { "grad_norm": 0.3016451299190521, "learning_rate": 2.9501544196038765e-05, "loss": 0.0193, "step": 52220 }, { "grad_norm": 0.2873188555240631, "learning_rate": 2.9482694337435614e-05, "loss": 0.0238, "step": 52230 }, { "grad_norm": 0.26805031299591064, "learning_rate": 2.946384798467887e-05, "loss": 0.0165, "step": 52240 }, { "grad_norm": 0.14122985303401947, "learning_rate": 2.9445005140988824e-05, "loss": 0.013, "step": 52250 }, { "grad_norm": 0.27010223269462585, "learning_rate": 2.942616580958524e-05, "loss": 0.0198, "step": 52260 }, { "grad_norm": 0.29852715134620667, "learning_rate": 2.9407329993687193e-05, "loss": 0.0174, "step": 52270 }, { "grad_norm": 0.2066735476255417, "learning_rate": 2.938849769651326e-05, "loss": 0.0157, "step": 52280 }, { "grad_norm": 0.2864101529121399, "learning_rate": 2.9369668921281345e-05, "loss": 0.0166, "step": 52290 }, { "grad_norm": 0.2541830539703369, "learning_rate": 2.9350843671208773e-05, "loss": 0.0138, "step": 52300 }, { "grad_norm": 0.25539806485176086, "learning_rate": 2.9332021949512245e-05, "loss": 0.0157, "step": 52310 }, { "grad_norm": 0.40905413031578064, "learning_rate": 2.9313203759407908e-05, "loss": 0.0199, "step": 52320 }, { "grad_norm": 0.1966654509305954, "learning_rate": 2.929438910411127e-05, "loss": 0.0294, "step": 52330 }, { "grad_norm": 0.2566298544406891, "learning_rate": 2.927557798683722e-05, "loss": 0.0169, "step": 52340 }, { "grad_norm": 0.19470258057117462, "learning_rate": 2.9256770410800095e-05, "loss": 0.0123, "step": 52350 }, { "grad_norm": 0.20236998796463013, "learning_rate": 2.9237966379213554e-05, "loss": 0.0178, "step": 52360 }, { "grad_norm": 0.22139491140842438, "learning_rate": 2.9219165895290736e-05, "loss": 0.0207, "step": 52370 }, { "grad_norm": 0.3582335114479065, "learning_rate": 2.9200368962244083e-05, "loss": 0.0201, "step": 52380 }, { "grad_norm": 0.22121846675872803, "learning_rate": 2.9181575583285536e-05, "loss": 0.0145, "step": 52390 }, { "grad_norm": 0.24068507552146912, "learning_rate": 2.916278576162632e-05, "loss": 0.0164, "step": 52400 }, { "grad_norm": 0.2562539875507355, "learning_rate": 2.9143999500477116e-05, "loss": 0.0163, "step": 52410 }, { "grad_norm": 0.2647715210914612, "learning_rate": 2.9125216803048004e-05, "loss": 0.021, "step": 52420 }, { "grad_norm": 0.19303864240646362, "learning_rate": 2.9106437672548403e-05, "loss": 0.0147, "step": 52430 }, { "grad_norm": 0.15430177748203278, "learning_rate": 2.9087662112187154e-05, "loss": 0.0126, "step": 52440 }, { "grad_norm": 0.2054930329322815, "learning_rate": 2.9068890125172498e-05, "loss": 0.0142, "step": 52450 }, { "grad_norm": 0.33745816349983215, "learning_rate": 2.9050121714712058e-05, "loss": 0.0163, "step": 52460 }, { "grad_norm": 0.32840925455093384, "learning_rate": 2.9031356884012835e-05, "loss": 0.0215, "step": 52470 }, { "grad_norm": 0.25843966007232666, "learning_rate": 2.9012595636281247e-05, "loss": 0.0189, "step": 52480 }, { "grad_norm": 0.5420458912849426, "learning_rate": 2.8993837974723016e-05, "loss": 0.0308, "step": 52490 }, { "grad_norm": 0.19595663249492645, "learning_rate": 2.8975083902543394e-05, "loss": 0.0175, "step": 52500 }, { "grad_norm": 0.3211478292942047, "learning_rate": 2.895633342294688e-05, "loss": 0.0192, "step": 52510 }, { "grad_norm": 0.21904459595680237, "learning_rate": 2.893758653913744e-05, "loss": 0.0186, "step": 52520 }, { "grad_norm": 0.19429633021354675, "learning_rate": 2.8918843254318423e-05, "loss": 0.017, "step": 52530 }, { "grad_norm": 0.2856045365333557, "learning_rate": 2.8900103571692483e-05, "loss": 0.0211, "step": 52540 }, { "grad_norm": 0.28316211700439453, "learning_rate": 2.8881367494461796e-05, "loss": 0.0169, "step": 52550 }, { "grad_norm": 0.19495904445648193, "learning_rate": 2.886263502582779e-05, "loss": 0.0151, "step": 52560 }, { "grad_norm": 0.260199636220932, "learning_rate": 2.8843906168991353e-05, "loss": 0.0164, "step": 52570 }, { "grad_norm": 0.2818760573863983, "learning_rate": 2.8825180927152728e-05, "loss": 0.0149, "step": 52580 }, { "grad_norm": 0.37438756227493286, "learning_rate": 2.8806459303511547e-05, "loss": 0.0148, "step": 52590 }, { "grad_norm": 0.27349209785461426, "learning_rate": 2.8787741301266852e-05, "loss": 0.0162, "step": 52600 }, { "grad_norm": 0.34112969040870667, "learning_rate": 2.876902692361697e-05, "loss": 0.0154, "step": 52610 }, { "grad_norm": 0.2478427290916443, "learning_rate": 2.875031617375976e-05, "loss": 0.0191, "step": 52620 }, { "grad_norm": 0.27554503083229065, "learning_rate": 2.8731609054892318e-05, "loss": 0.0167, "step": 52630 }, { "grad_norm": 0.1901194304227829, "learning_rate": 2.871290557021119e-05, "loss": 0.0154, "step": 52640 }, { "grad_norm": 0.18000566959381104, "learning_rate": 2.8694205722912298e-05, "loss": 0.0209, "step": 52650 }, { "grad_norm": 0.29586315155029297, "learning_rate": 2.8675509516190936e-05, "loss": 0.02, "step": 52660 }, { "grad_norm": 0.23998823761940002, "learning_rate": 2.8656816953241772e-05, "loss": 0.0203, "step": 52670 }, { "grad_norm": 0.28736311197280884, "learning_rate": 2.863812803725887e-05, "loss": 0.0142, "step": 52680 }, { "grad_norm": 0.308354914188385, "learning_rate": 2.8619442771435623e-05, "loss": 0.0168, "step": 52690 }, { "grad_norm": 0.15344049036502838, "learning_rate": 2.860076115896484e-05, "loss": 0.0196, "step": 52700 }, { "grad_norm": 0.18511468172073364, "learning_rate": 2.8582083203038712e-05, "loss": 0.0168, "step": 52710 }, { "grad_norm": 0.1985727846622467, "learning_rate": 2.8563408906848778e-05, "loss": 0.015, "step": 52720 }, { "grad_norm": 0.37515389919281006, "learning_rate": 2.8544738273585993e-05, "loss": 0.0174, "step": 52730 }, { "grad_norm": 0.23236975073814392, "learning_rate": 2.8526071306440595e-05, "loss": 0.0148, "step": 52740 }, { "grad_norm": 0.2916312515735626, "learning_rate": 2.850740800860233e-05, "loss": 0.0131, "step": 52750 }, { "grad_norm": 0.17636461555957794, "learning_rate": 2.8488748383260178e-05, "loss": 0.0149, "step": 52760 }, { "grad_norm": 0.21566101908683777, "learning_rate": 2.8470092433602635e-05, "loss": 0.0222, "step": 52770 }, { "grad_norm": 0.30820149183273315, "learning_rate": 2.845144016281745e-05, "loss": 0.0147, "step": 52780 }, { "grad_norm": 0.3171870708465576, "learning_rate": 2.8432791574091744e-05, "loss": 0.0292, "step": 52790 }, { "grad_norm": 0.3083791136741638, "learning_rate": 2.8414146670612134e-05, "loss": 0.022, "step": 52800 }, { "grad_norm": 0.20248793065547943, "learning_rate": 2.8395505455564446e-05, "loss": 0.0153, "step": 52810 }, { "grad_norm": 0.23025335371494293, "learning_rate": 2.837686793213403e-05, "loss": 0.0145, "step": 52820 }, { "grad_norm": 0.3072664439678192, "learning_rate": 2.8358234103505477e-05, "loss": 0.0148, "step": 52830 }, { "grad_norm": 0.20204509794712067, "learning_rate": 2.8339603972862806e-05, "loss": 0.0155, "step": 52840 }, { "grad_norm": 0.2434028685092926, "learning_rate": 2.83209775433894e-05, "loss": 0.019, "step": 52850 }, { "grad_norm": 0.4125862419605255, "learning_rate": 2.8302354818268013e-05, "loss": 0.0186, "step": 52860 }, { "grad_norm": 0.29255250096321106, "learning_rate": 2.8283735800680754e-05, "loss": 0.017, "step": 52870 }, { "grad_norm": 0.3120977580547333, "learning_rate": 2.826512049380913e-05, "loss": 0.0208, "step": 52880 }, { "grad_norm": 0.22521457076072693, "learning_rate": 2.8246508900833935e-05, "loss": 0.0189, "step": 52890 }, { "grad_norm": 0.2636635899543762, "learning_rate": 2.822790102493541e-05, "loss": 0.0212, "step": 52900 }, { "grad_norm": 0.2255949229001999, "learning_rate": 2.820929686929314e-05, "loss": 0.0161, "step": 52910 }, { "grad_norm": 0.2516242265701294, "learning_rate": 2.819069643708605e-05, "loss": 0.0161, "step": 52920 }, { "grad_norm": 0.4027464687824249, "learning_rate": 2.8172099731492484e-05, "loss": 0.0179, "step": 52930 }, { "grad_norm": 0.26797497272491455, "learning_rate": 2.815350675569003e-05, "loss": 0.0238, "step": 52940 }, { "grad_norm": 0.23101742565631866, "learning_rate": 2.8134917512855807e-05, "loss": 0.0208, "step": 52950 }, { "grad_norm": 0.17031164467334747, "learning_rate": 2.8116332006166134e-05, "loss": 0.0175, "step": 52960 }, { "grad_norm": 0.3088986277580261, "learning_rate": 2.809775023879685e-05, "loss": 0.024, "step": 52970 }, { "grad_norm": 0.2642349898815155, "learning_rate": 2.807917221392299e-05, "loss": 0.0188, "step": 52980 }, { "grad_norm": 0.20840774476528168, "learning_rate": 2.8060597934719067e-05, "loss": 0.0179, "step": 52990 }, { "grad_norm": 0.43853676319122314, "learning_rate": 2.8042027404358927e-05, "loss": 0.0163, "step": 53000 }, { "grad_norm": 0.24367882311344147, "learning_rate": 2.802346062601572e-05, "loss": 0.0174, "step": 53010 }, { "grad_norm": 0.38183531165122986, "learning_rate": 2.800489760286207e-05, "loss": 0.0218, "step": 53020 }, { "grad_norm": 0.27747347950935364, "learning_rate": 2.7986338338069834e-05, "loss": 0.018, "step": 53030 }, { "grad_norm": 0.35458144545555115, "learning_rate": 2.7967782834810297e-05, "loss": 0.0212, "step": 53040 }, { "grad_norm": 0.22184649109840393, "learning_rate": 2.794923109625409e-05, "loss": 0.0117, "step": 53050 }, { "grad_norm": 0.3450097143650055, "learning_rate": 2.7930683125571205e-05, "loss": 0.0163, "step": 53060 }, { "grad_norm": 0.13233080506324768, "learning_rate": 2.7912138925930997e-05, "loss": 0.0201, "step": 53070 }, { "grad_norm": 0.23498228192329407, "learning_rate": 2.7893598500502117e-05, "loss": 0.0265, "step": 53080 }, { "grad_norm": 0.33796799182891846, "learning_rate": 2.7875061852452644e-05, "loss": 0.0236, "step": 53090 }, { "grad_norm": 0.3665207624435425, "learning_rate": 2.7856528984949982e-05, "loss": 0.0169, "step": 53100 }, { "grad_norm": 0.20877854526042938, "learning_rate": 2.7837999901160888e-05, "loss": 0.0191, "step": 53110 }, { "grad_norm": 0.20915135741233826, "learning_rate": 2.7819474604251484e-05, "loss": 0.015, "step": 53120 }, { "grad_norm": 0.260195791721344, "learning_rate": 2.780095309738725e-05, "loss": 0.0193, "step": 53130 }, { "grad_norm": 0.427409291267395, "learning_rate": 2.778243538373294e-05, "loss": 0.0164, "step": 53140 }, { "grad_norm": 0.2141939401626587, "learning_rate": 2.7763921466452826e-05, "loss": 0.018, "step": 53150 }, { "grad_norm": 0.3245006501674652, "learning_rate": 2.7745411348710336e-05, "loss": 0.0235, "step": 53160 }, { "grad_norm": 0.541252851486206, "learning_rate": 2.7726905033668395e-05, "loss": 0.025, "step": 53170 }, { "grad_norm": 0.37753528356552124, "learning_rate": 2.7708402524489214e-05, "loss": 0.0203, "step": 53180 }, { "grad_norm": 0.2438202202320099, "learning_rate": 2.7689903824334364e-05, "loss": 0.0234, "step": 53190 }, { "grad_norm": 0.28711843490600586, "learning_rate": 2.7671408936364785e-05, "loss": 0.016, "step": 53200 }, { "grad_norm": 0.34781181812286377, "learning_rate": 2.765291786374069e-05, "loss": 0.0145, "step": 53210 }, { "grad_norm": 0.1518811285495758, "learning_rate": 2.7634430609621775e-05, "loss": 0.0132, "step": 53220 }, { "grad_norm": 0.6724228262901306, "learning_rate": 2.7615947177166956e-05, "loss": 0.019, "step": 53230 }, { "grad_norm": 0.2264455109834671, "learning_rate": 2.7597467569534553e-05, "loss": 0.0146, "step": 53240 }, { "grad_norm": 0.31123656034469604, "learning_rate": 2.757899178988226e-05, "loss": 0.0174, "step": 53250 }, { "grad_norm": 0.33295246958732605, "learning_rate": 2.7560519841367005e-05, "loss": 0.0148, "step": 53260 }, { "grad_norm": 0.22664770483970642, "learning_rate": 2.7542051727145237e-05, "loss": 0.0179, "step": 53270 }, { "grad_norm": 0.2685011923313141, "learning_rate": 2.7523587450372578e-05, "loss": 0.0224, "step": 53280 }, { "grad_norm": 0.35075369477272034, "learning_rate": 2.750512701420409e-05, "loss": 0.0157, "step": 53290 }, { "grad_norm": 0.3490120470523834, "learning_rate": 2.7486670421794158e-05, "loss": 0.0203, "step": 53300 }, { "grad_norm": 0.26261189579963684, "learning_rate": 2.7468217676296515e-05, "loss": 0.0266, "step": 53310 }, { "grad_norm": 0.3128761053085327, "learning_rate": 2.7449768780864226e-05, "loss": 0.0204, "step": 53320 }, { "grad_norm": 0.1984582394361496, "learning_rate": 2.7431323738649724e-05, "loss": 0.0168, "step": 53330 }, { "grad_norm": 0.3160400390625, "learning_rate": 2.7412882552804713e-05, "loss": 0.0185, "step": 53340 }, { "grad_norm": 0.33288925886154175, "learning_rate": 2.739444522648032e-05, "loss": 0.0142, "step": 53350 }, { "grad_norm": 0.28115835785865784, "learning_rate": 2.737601176282697e-05, "loss": 0.0199, "step": 53360 }, { "grad_norm": 0.45093101263046265, "learning_rate": 2.735758216499445e-05, "loss": 0.0152, "step": 53370 }, { "grad_norm": 0.2585451006889343, "learning_rate": 2.7339156436131864e-05, "loss": 0.0155, "step": 53380 }, { "grad_norm": 0.29426300525665283, "learning_rate": 2.7320734579387663e-05, "loss": 0.0179, "step": 53390 }, { "grad_norm": 0.2619844973087311, "learning_rate": 2.730231659790966e-05, "loss": 0.0139, "step": 53400 }, { "grad_norm": 0.2219667136669159, "learning_rate": 2.728390249484494e-05, "loss": 0.0217, "step": 53410 }, { "grad_norm": 0.2635265290737152, "learning_rate": 2.7265492273340032e-05, "loss": 0.0176, "step": 53420 }, { "grad_norm": 0.2544519603252411, "learning_rate": 2.7247085936540678e-05, "loss": 0.0149, "step": 53430 }, { "grad_norm": 0.22836928069591522, "learning_rate": 2.722868348759205e-05, "loss": 0.0178, "step": 53440 }, { "grad_norm": 0.33827921748161316, "learning_rate": 2.7210284929638635e-05, "loss": 0.02, "step": 53450 }, { "grad_norm": 0.32640424370765686, "learning_rate": 2.7191890265824183e-05, "loss": 0.0157, "step": 53460 }, { "grad_norm": 0.1992044299840927, "learning_rate": 2.7173499499291926e-05, "loss": 0.0137, "step": 53470 }, { "grad_norm": 0.3479847013950348, "learning_rate": 2.7155112633184277e-05, "loss": 0.014, "step": 53480 }, { "grad_norm": 0.283138632774353, "learning_rate": 2.713672967064307e-05, "loss": 0.0237, "step": 53490 }, { "grad_norm": 0.2884293794631958, "learning_rate": 2.711835061480945e-05, "loss": 0.013, "step": 53500 }, { "grad_norm": 0.29186517000198364, "learning_rate": 2.7099975468823896e-05, "loss": 0.0237, "step": 53510 }, { "grad_norm": 0.3145194947719574, "learning_rate": 2.708160423582622e-05, "loss": 0.0166, "step": 53520 }, { "grad_norm": 0.5480833053588867, "learning_rate": 2.706323691895557e-05, "loss": 0.0195, "step": 53530 }, { "grad_norm": 0.2977868914604187, "learning_rate": 2.70448735213504e-05, "loss": 0.0169, "step": 53540 }, { "grad_norm": 0.2087927609682083, "learning_rate": 2.702651404614852e-05, "loss": 0.017, "step": 53550 }, { "grad_norm": 0.21095393598079681, "learning_rate": 2.7008158496487056e-05, "loss": 0.0147, "step": 53560 }, { "grad_norm": 0.23144367337226868, "learning_rate": 2.6989806875502487e-05, "loss": 0.0151, "step": 53570 }, { "grad_norm": 0.2511337101459503, "learning_rate": 2.6971459186330584e-05, "loss": 0.0127, "step": 53580 }, { "grad_norm": 0.2854858636856079, "learning_rate": 2.695311543210648e-05, "loss": 0.0164, "step": 53590 }, { "grad_norm": 0.2551407516002655, "learning_rate": 2.693477561596463e-05, "loss": 0.0156, "step": 53600 }, { "grad_norm": 0.27481895685195923, "learning_rate": 2.6916439741038756e-05, "loss": 0.0166, "step": 53610 }, { "grad_norm": 0.36374619603157043, "learning_rate": 2.689810781046203e-05, "loss": 0.0192, "step": 53620 }, { "grad_norm": 0.284183144569397, "learning_rate": 2.6879779827366823e-05, "loss": 0.0174, "step": 53630 }, { "grad_norm": 0.23008789122104645, "learning_rate": 2.6861455794884904e-05, "loss": 0.018, "step": 53640 }, { "grad_norm": 0.22390271723270416, "learning_rate": 2.6843135716147373e-05, "loss": 0.0161, "step": 53650 }, { "grad_norm": 0.23822174966335297, "learning_rate": 2.6824819594284556e-05, "loss": 0.0187, "step": 53660 }, { "grad_norm": 0.28970763087272644, "learning_rate": 2.6806507432426275e-05, "loss": 0.0155, "step": 53670 }, { "grad_norm": 0.3679524064064026, "learning_rate": 2.6788199233701512e-05, "loss": 0.018, "step": 53680 }, { "grad_norm": 0.28989458084106445, "learning_rate": 2.6769895001238652e-05, "loss": 0.0191, "step": 53690 }, { "grad_norm": 0.25871580839157104, "learning_rate": 2.67515947381654e-05, "loss": 0.0158, "step": 53700 }, { "grad_norm": 0.14568521082401276, "learning_rate": 2.6733298447608745e-05, "loss": 0.0154, "step": 53710 }, { "grad_norm": 0.18737973272800446, "learning_rate": 2.6715006132695074e-05, "loss": 0.0209, "step": 53720 }, { "grad_norm": 0.2291259467601776, "learning_rate": 2.6696717796549985e-05, "loss": 0.023, "step": 53730 }, { "grad_norm": 0.340791791677475, "learning_rate": 2.667843344229848e-05, "loss": 0.0182, "step": 53740 }, { "grad_norm": 0.4218696355819702, "learning_rate": 2.6660153073064846e-05, "loss": 0.0161, "step": 53750 }, { "grad_norm": 0.2917448580265045, "learning_rate": 2.664187669197271e-05, "loss": 0.03, "step": 53760 }, { "grad_norm": 0.26285678148269653, "learning_rate": 2.6623604302145005e-05, "loss": 0.0167, "step": 53770 }, { "grad_norm": 0.3778233230113983, "learning_rate": 2.6605335906703975e-05, "loss": 0.0206, "step": 53780 }, { "grad_norm": 0.1734369397163391, "learning_rate": 2.65870715087712e-05, "loss": 0.0181, "step": 53790 }, { "grad_norm": 0.31685367226600647, "learning_rate": 2.6568811111467573e-05, "loss": 0.0128, "step": 53800 }, { "grad_norm": 0.16581936180591583, "learning_rate": 2.6550554717913258e-05, "loss": 0.0171, "step": 53810 }, { "grad_norm": 0.2350476086139679, "learning_rate": 2.6532302331227805e-05, "loss": 0.0154, "step": 53820 }, { "grad_norm": 0.41005128622055054, "learning_rate": 2.651405395453004e-05, "loss": 0.0116, "step": 53830 }, { "grad_norm": 0.17940035462379456, "learning_rate": 2.6495809590938115e-05, "loss": 0.0244, "step": 53840 }, { "grad_norm": 0.9021288752555847, "learning_rate": 2.64775692435695e-05, "loss": 0.0158, "step": 53850 }, { "grad_norm": 0.25029662251472473, "learning_rate": 2.6459332915540928e-05, "loss": 0.0183, "step": 53860 }, { "grad_norm": 0.24339492619037628, "learning_rate": 2.644110060996856e-05, "loss": 0.0244, "step": 53870 }, { "grad_norm": 0.21192742884159088, "learning_rate": 2.642287232996774e-05, "loss": 0.0137, "step": 53880 }, { "grad_norm": 0.195151686668396, "learning_rate": 2.6404648078653205e-05, "loss": 0.0117, "step": 53890 }, { "grad_norm": 0.35937392711639404, "learning_rate": 2.6386427859139002e-05, "loss": 0.0151, "step": 53900 }, { "grad_norm": 0.49202731251716614, "learning_rate": 2.63682116745384e-05, "loss": 0.023, "step": 53910 }, { "grad_norm": 0.2740524709224701, "learning_rate": 2.6349999527964138e-05, "loss": 0.0163, "step": 53920 }, { "grad_norm": 0.3435159921646118, "learning_rate": 2.63317914225281e-05, "loss": 0.0235, "step": 53930 }, { "grad_norm": 0.2958311438560486, "learning_rate": 2.6313587361341585e-05, "loss": 0.018, "step": 53940 }, { "grad_norm": 0.422882616519928, "learning_rate": 2.6295387347515165e-05, "loss": 0.0252, "step": 53950 }, { "grad_norm": 0.1881960779428482, "learning_rate": 2.6277191384158727e-05, "loss": 0.0201, "step": 53960 }, { "grad_norm": 0.2929237186908722, "learning_rate": 2.6258999474381457e-05, "loss": 0.0194, "step": 53970 }, { "grad_norm": 0.28263676166534424, "learning_rate": 2.624081162129186e-05, "loss": 0.014, "step": 53980 }, { "grad_norm": 0.27013665437698364, "learning_rate": 2.6222627827997765e-05, "loss": 0.0257, "step": 53990 }, { "grad_norm": 0.4853264391422272, "learning_rate": 2.6204448097606236e-05, "loss": 0.0165, "step": 54000 }, { "grad_norm": 0.2470589578151703, "learning_rate": 2.6186272433223726e-05, "loss": 0.019, "step": 54010 }, { "grad_norm": 0.23376983404159546, "learning_rate": 2.6168100837955943e-05, "loss": 0.0178, "step": 54020 }, { "grad_norm": 0.37026408314704895, "learning_rate": 2.6149933314907926e-05, "loss": 0.0256, "step": 54030 }, { "grad_norm": 0.3924727141857147, "learning_rate": 2.6131769867184e-05, "loss": 0.0253, "step": 54040 }, { "grad_norm": 0.2463269978761673, "learning_rate": 2.611361049788783e-05, "loss": 0.0164, "step": 54050 }, { "grad_norm": 0.388790100812912, "learning_rate": 2.6095455210122292e-05, "loss": 0.0165, "step": 54060 }, { "grad_norm": 0.23833882808685303, "learning_rate": 2.6077304006989712e-05, "loss": 0.0141, "step": 54070 }, { "grad_norm": 0.2857954800128937, "learning_rate": 2.6059156891591562e-05, "loss": 0.0198, "step": 54080 }, { "grad_norm": 0.31212151050567627, "learning_rate": 2.6041013867028718e-05, "loss": 0.0163, "step": 54090 }, { "grad_norm": 0.40667733550071716, "learning_rate": 2.6022874936401347e-05, "loss": 0.0157, "step": 54100 }, { "grad_norm": 0.15343131124973297, "learning_rate": 2.6004740102808832e-05, "loss": 0.0194, "step": 54110 }, { "grad_norm": 0.24377848207950592, "learning_rate": 2.598660936935e-05, "loss": 0.0162, "step": 54120 }, { "grad_norm": 0.2787661850452423, "learning_rate": 2.5968482739122845e-05, "loss": 0.0219, "step": 54130 }, { "grad_norm": 0.30337104201316833, "learning_rate": 2.595036021522472e-05, "loss": 0.0195, "step": 54140 }, { "grad_norm": 0.31486424803733826, "learning_rate": 2.5932241800752278e-05, "loss": 0.0154, "step": 54150 }, { "grad_norm": 0.3423336446285248, "learning_rate": 2.5914127498801453e-05, "loss": 0.0272, "step": 54160 }, { "grad_norm": 0.22619961202144623, "learning_rate": 2.5896017312467497e-05, "loss": 0.0199, "step": 54170 }, { "grad_norm": 0.26661860942840576, "learning_rate": 2.587791124484493e-05, "loss": 0.0123, "step": 54180 }, { "grad_norm": 0.13828569650650024, "learning_rate": 2.5859809299027615e-05, "loss": 0.0182, "step": 54190 }, { "grad_norm": 0.2329953908920288, "learning_rate": 2.5841711478108632e-05, "loss": 0.0182, "step": 54200 }, { "grad_norm": 0.325717031955719, "learning_rate": 2.582361778518043e-05, "loss": 0.019, "step": 54210 }, { "grad_norm": 0.4945259094238281, "learning_rate": 2.580552822333472e-05, "loss": 0.0214, "step": 54220 }, { "grad_norm": 0.26606515049934387, "learning_rate": 2.578744279566252e-05, "loss": 0.0141, "step": 54230 }, { "grad_norm": 0.2940373420715332, "learning_rate": 2.576936150525413e-05, "loss": 0.0219, "step": 54240 }, { "grad_norm": 0.13754084706306458, "learning_rate": 2.5751284355199168e-05, "loss": 0.0139, "step": 54250 }, { "grad_norm": 0.48132771253585815, "learning_rate": 2.573321134858646e-05, "loss": 0.0166, "step": 54260 }, { "grad_norm": 0.2224649339914322, "learning_rate": 2.5715142488504286e-05, "loss": 0.0136, "step": 54270 }, { "grad_norm": 0.2277238517999649, "learning_rate": 2.5697077778040042e-05, "loss": 0.0176, "step": 54280 }, { "grad_norm": 0.2151923030614853, "learning_rate": 2.5679017220280522e-05, "loss": 0.0272, "step": 54290 }, { "grad_norm": 0.19661976397037506, "learning_rate": 2.5660960818311796e-05, "loss": 0.018, "step": 54300 }, { "grad_norm": 0.39271536469459534, "learning_rate": 2.564290857521915e-05, "loss": 0.0195, "step": 54310 }, { "grad_norm": 0.27582141757011414, "learning_rate": 2.5624860494087298e-05, "loss": 0.0172, "step": 54320 }, { "grad_norm": 0.2578660547733307, "learning_rate": 2.5606816578000115e-05, "loss": 0.0207, "step": 54330 }, { "grad_norm": 0.3315611481666565, "learning_rate": 2.558877683004082e-05, "loss": 0.0156, "step": 54340 }, { "grad_norm": 0.2202124297618866, "learning_rate": 2.557074125329192e-05, "loss": 0.0153, "step": 54350 }, { "grad_norm": 0.18885643780231476, "learning_rate": 2.5552709850835195e-05, "loss": 0.0161, "step": 54360 }, { "grad_norm": 0.15254440903663635, "learning_rate": 2.5534682625751738e-05, "loss": 0.0139, "step": 54370 }, { "grad_norm": 0.177597314119339, "learning_rate": 2.551665958112186e-05, "loss": 0.0266, "step": 54380 }, { "grad_norm": 0.4273591935634613, "learning_rate": 2.549864072002527e-05, "loss": 0.0151, "step": 54390 }, { "grad_norm": 0.27140679955482483, "learning_rate": 2.5480626045540858e-05, "loss": 0.0165, "step": 54400 }, { "grad_norm": 0.2717150151729584, "learning_rate": 2.546261556074684e-05, "loss": 0.0155, "step": 54410 }, { "grad_norm": 0.26927274465560913, "learning_rate": 2.5444609268720726e-05, "loss": 0.0158, "step": 54420 }, { "grad_norm": 0.2414720505475998, "learning_rate": 2.5426607172539297e-05, "loss": 0.0155, "step": 54430 }, { "grad_norm": 0.34942853450775146, "learning_rate": 2.5408609275278617e-05, "loss": 0.0259, "step": 54440 }, { "grad_norm": 0.18058288097381592, "learning_rate": 2.5390615580014055e-05, "loss": 0.0146, "step": 54450 }, { "grad_norm": 0.20739015936851501, "learning_rate": 2.5372626089820207e-05, "loss": 0.0214, "step": 54460 }, { "grad_norm": 0.2595542371273041, "learning_rate": 2.5354640807770997e-05, "loss": 0.0132, "step": 54470 }, { "grad_norm": 0.31119489669799805, "learning_rate": 2.5336659736939622e-05, "loss": 0.0201, "step": 54480 }, { "grad_norm": 0.3117067515850067, "learning_rate": 2.5318682880398554e-05, "loss": 0.0182, "step": 54490 }, { "grad_norm": 0.24634695053100586, "learning_rate": 2.530071024121956e-05, "loss": 0.0147, "step": 54500 }, { "grad_norm": 0.25253191590309143, "learning_rate": 2.5282741822473627e-05, "loss": 0.0151, "step": 54510 }, { "grad_norm": 0.3323051929473877, "learning_rate": 2.526477762723114e-05, "loss": 0.0174, "step": 54520 }, { "grad_norm": 0.1578199416399002, "learning_rate": 2.5246817658561618e-05, "loss": 0.0168, "step": 54530 }, { "grad_norm": 0.21884138882160187, "learning_rate": 2.5228861919533965e-05, "loss": 0.0151, "step": 54540 }, { "grad_norm": 0.18549717962741852, "learning_rate": 2.5210910413216326e-05, "loss": 0.0276, "step": 54550 }, { "grad_norm": 0.21067756414413452, "learning_rate": 2.5192963142676086e-05, "loss": 0.0145, "step": 54560 }, { "grad_norm": 0.2506959140300751, "learning_rate": 2.517502011098001e-05, "loss": 0.016, "step": 54570 }, { "grad_norm": 0.369316965341568, "learning_rate": 2.5157081321193987e-05, "loss": 0.0193, "step": 54580 }, { "grad_norm": 0.11985871940851212, "learning_rate": 2.5139146776383356e-05, "loss": 0.0152, "step": 54590 }, { "grad_norm": 0.22043465077877045, "learning_rate": 2.5121216479612575e-05, "loss": 0.0142, "step": 54600 }, { "grad_norm": 0.21274229884147644, "learning_rate": 2.510329043394546e-05, "loss": 0.02, "step": 54610 }, { "grad_norm": 0.4857528507709503, "learning_rate": 2.508536864244508e-05, "loss": 0.0198, "step": 54620 }, { "grad_norm": 0.23728860914707184, "learning_rate": 2.5067451108173778e-05, "loss": 0.0149, "step": 54630 }, { "grad_norm": 0.2389395534992218, "learning_rate": 2.5049537834193204e-05, "loss": 0.0113, "step": 54640 }, { "grad_norm": 0.170305535197258, "learning_rate": 2.5031628823564194e-05, "loss": 0.0172, "step": 54650 }, { "grad_norm": 0.3478405177593231, "learning_rate": 2.5013724079346933e-05, "loss": 0.0164, "step": 54660 }, { "grad_norm": 0.3893091678619385, "learning_rate": 2.4995823604600854e-05, "loss": 0.0129, "step": 54670 }, { "grad_norm": 0.321439266204834, "learning_rate": 2.497792740238465e-05, "loss": 0.0165, "step": 54680 }, { "grad_norm": 0.28748375177383423, "learning_rate": 2.49600354757563e-05, "loss": 0.0227, "step": 54690 }, { "grad_norm": 0.21923959255218506, "learning_rate": 2.494214782777306e-05, "loss": 0.0211, "step": 54700 }, { "grad_norm": 0.3579785227775574, "learning_rate": 2.4924264461491386e-05, "loss": 0.0175, "step": 54710 }, { "grad_norm": 0.3700982332229614, "learning_rate": 2.4906385379967133e-05, "loss": 0.0211, "step": 54720 }, { "grad_norm": 0.2510780096054077, "learning_rate": 2.4888510586255285e-05, "loss": 0.0191, "step": 54730 }, { "grad_norm": 0.2266823649406433, "learning_rate": 2.487064008341018e-05, "loss": 0.0173, "step": 54740 }, { "grad_norm": 0.32275718450546265, "learning_rate": 2.4852773874485407e-05, "loss": 0.0153, "step": 54750 }, { "grad_norm": 0.27939537167549133, "learning_rate": 2.483491196253377e-05, "loss": 0.019, "step": 54760 }, { "grad_norm": 0.49740538001060486, "learning_rate": 2.4817054350607443e-05, "loss": 0.0222, "step": 54770 }, { "grad_norm": 0.23791125416755676, "learning_rate": 2.4799201041757742e-05, "loss": 0.0185, "step": 54780 }, { "grad_norm": 0.3340466320514679, "learning_rate": 2.4781352039035373e-05, "loss": 0.0179, "step": 54790 }, { "grad_norm": 0.24413742125034332, "learning_rate": 2.4763507345490194e-05, "loss": 0.018, "step": 54800 }, { "grad_norm": 0.35816025733947754, "learning_rate": 2.4745666964171386e-05, "loss": 0.0227, "step": 54810 }, { "grad_norm": 0.386120468378067, "learning_rate": 2.4727830898127387e-05, "loss": 0.0242, "step": 54820 }, { "grad_norm": 0.27723070979118347, "learning_rate": 2.4709999150405895e-05, "loss": 0.0163, "step": 54830 }, { "grad_norm": 0.4063574969768524, "learning_rate": 2.469217172405388e-05, "loss": 0.0191, "step": 54840 }, { "grad_norm": 0.18954189121723175, "learning_rate": 2.4674348622117527e-05, "loss": 0.0122, "step": 54850 }, { "grad_norm": 0.18480902910232544, "learning_rate": 2.465652984764234e-05, "loss": 0.0147, "step": 54860 }, { "grad_norm": 0.24712419509887695, "learning_rate": 2.4638715403673056e-05, "loss": 0.0141, "step": 54870 }, { "grad_norm": 0.29341790080070496, "learning_rate": 2.4620905293253676e-05, "loss": 0.0207, "step": 54880 }, { "grad_norm": 0.19944605231285095, "learning_rate": 2.4603099519427463e-05, "loss": 0.0168, "step": 54890 }, { "grad_norm": 0.2810033857822418, "learning_rate": 2.458529808523695e-05, "loss": 0.0188, "step": 54900 }, { "grad_norm": 0.2283448576927185, "learning_rate": 2.456750099372387e-05, "loss": 0.0145, "step": 54910 }, { "grad_norm": 0.1826609969139099, "learning_rate": 2.4549708247929327e-05, "loss": 0.0188, "step": 54920 }, { "grad_norm": 0.24039746820926666, "learning_rate": 2.4531919850893554e-05, "loss": 0.0124, "step": 54930 }, { "grad_norm": 0.19887830317020416, "learning_rate": 2.4514135805656125e-05, "loss": 0.0135, "step": 54940 }, { "grad_norm": 0.4207056164741516, "learning_rate": 2.449635611525587e-05, "loss": 0.0201, "step": 54950 }, { "grad_norm": 0.2941076457500458, "learning_rate": 2.447858078273079e-05, "loss": 0.0172, "step": 54960 }, { "grad_norm": 0.3595753610134125, "learning_rate": 2.4460809811118275e-05, "loss": 0.0331, "step": 54970 }, { "grad_norm": 0.25610193610191345, "learning_rate": 2.444304320345483e-05, "loss": 0.0181, "step": 54980 }, { "grad_norm": 0.31324952840805054, "learning_rate": 2.4425280962776354e-05, "loss": 0.0215, "step": 54990 }, { "grad_norm": 0.2727285921573639, "learning_rate": 2.4407523092117875e-05, "loss": 0.016, "step": 55000 }, { "grad_norm": 0.35195690393447876, "learning_rate": 2.438976959451374e-05, "loss": 0.0183, "step": 55010 }, { "grad_norm": 0.2953406870365143, "learning_rate": 2.4372020472997565e-05, "loss": 0.0156, "step": 55020 }, { "grad_norm": 0.2836606502532959, "learning_rate": 2.4354275730602122e-05, "loss": 0.0188, "step": 55030 }, { "grad_norm": 0.26590725779533386, "learning_rate": 2.433653537035958e-05, "loss": 0.0217, "step": 55040 }, { "grad_norm": 0.25885477662086487, "learning_rate": 2.431879939530123e-05, "loss": 0.0169, "step": 55050 }, { "grad_norm": 0.3438626527786255, "learning_rate": 2.4301067808457684e-05, "loss": 0.0166, "step": 55060 }, { "grad_norm": 0.2589937150478363, "learning_rate": 2.428334061285878e-05, "loss": 0.0149, "step": 55070 }, { "grad_norm": 0.2980611026287079, "learning_rate": 2.426561781153361e-05, "loss": 0.0161, "step": 55080 }, { "grad_norm": 0.2358178049325943, "learning_rate": 2.424789940751052e-05, "loss": 0.0173, "step": 55090 }, { "grad_norm": 0.2205834835767746, "learning_rate": 2.423018540381712e-05, "loss": 0.0157, "step": 55100 }, { "grad_norm": 0.12795431911945343, "learning_rate": 2.421247580348021e-05, "loss": 0.0157, "step": 55110 }, { "grad_norm": 0.25161686539649963, "learning_rate": 2.4194770609525897e-05, "loss": 0.0151, "step": 55120 }, { "grad_norm": 0.23641082644462585, "learning_rate": 2.417706982497951e-05, "loss": 0.0181, "step": 55130 }, { "grad_norm": 0.2580108940601349, "learning_rate": 2.415937345286563e-05, "loss": 0.0171, "step": 55140 }, { "grad_norm": 0.26832377910614014, "learning_rate": 2.4141681496208087e-05, "loss": 0.0129, "step": 55150 }, { "grad_norm": 0.21249769628047943, "learning_rate": 2.4123993958029946e-05, "loss": 0.0142, "step": 55160 }, { "grad_norm": 0.23509478569030762, "learning_rate": 2.4106310841353548e-05, "loss": 0.0148, "step": 55170 }, { "grad_norm": 0.1919584721326828, "learning_rate": 2.4088632149200398e-05, "loss": 0.0172, "step": 55180 }, { "grad_norm": 0.2058178186416626, "learning_rate": 2.4070957884591367e-05, "loss": 0.0128, "step": 55190 }, { "grad_norm": 0.5257717967033386, "learning_rate": 2.4053288050546464e-05, "loss": 0.0219, "step": 55200 }, { "grad_norm": 0.334539532661438, "learning_rate": 2.403562265008498e-05, "loss": 0.0176, "step": 55210 }, { "grad_norm": 0.2981782555580139, "learning_rate": 2.4017961686225483e-05, "loss": 0.018, "step": 55220 }, { "grad_norm": 0.2602614760398865, "learning_rate": 2.400030516198568e-05, "loss": 0.0219, "step": 55230 }, { "grad_norm": 0.2346372902393341, "learning_rate": 2.3982653080382673e-05, "loss": 0.0141, "step": 55240 }, { "grad_norm": 0.159299835562706, "learning_rate": 2.396500544443266e-05, "loss": 0.0177, "step": 55250 }, { "grad_norm": 0.35492846369743347, "learning_rate": 2.3947362257151156e-05, "loss": 0.0163, "step": 55260 }, { "grad_norm": 0.36486509442329407, "learning_rate": 2.39297235215529e-05, "loss": 0.0235, "step": 55270 }, { "grad_norm": 0.26811766624450684, "learning_rate": 2.3912089240651873e-05, "loss": 0.019, "step": 55280 }, { "grad_norm": 0.29867905378341675, "learning_rate": 2.389445941746129e-05, "loss": 0.0176, "step": 55290 }, { "grad_norm": 0.2503436803817749, "learning_rate": 2.3876834054993625e-05, "loss": 0.0144, "step": 55300 }, { "grad_norm": 0.22273202240467072, "learning_rate": 2.3859213156260522e-05, "loss": 0.0166, "step": 55310 }, { "grad_norm": 0.4331844449043274, "learning_rate": 2.3841596724272948e-05, "loss": 0.0187, "step": 55320 }, { "grad_norm": 0.2576422393321991, "learning_rate": 2.382398476204106e-05, "loss": 0.015, "step": 55330 }, { "grad_norm": 0.19272354245185852, "learning_rate": 2.3806377272574254e-05, "loss": 0.0132, "step": 55340 }, { "grad_norm": 0.5068474411964417, "learning_rate": 2.3788774258881174e-05, "loss": 0.0179, "step": 55350 }, { "grad_norm": 0.2557965815067291, "learning_rate": 2.37711757239697e-05, "loss": 0.019, "step": 55360 }, { "grad_norm": 0.4403766393661499, "learning_rate": 2.3753581670846954e-05, "loss": 0.0165, "step": 55370 }, { "grad_norm": 0.19931857287883759, "learning_rate": 2.3735992102519216e-05, "loss": 0.0173, "step": 55380 }, { "grad_norm": 0.24628953635692596, "learning_rate": 2.371840702199215e-05, "loss": 0.0154, "step": 55390 }, { "grad_norm": 0.25850653648376465, "learning_rate": 2.3700826432270494e-05, "loss": 0.0147, "step": 55400 }, { "grad_norm": 0.31364044547080994, "learning_rate": 2.3683250336358326e-05, "loss": 0.024, "step": 55410 }, { "grad_norm": 0.20715749263763428, "learning_rate": 2.3665678737258923e-05, "loss": 0.0182, "step": 55420 }, { "grad_norm": 0.11567170172929764, "learning_rate": 2.3648111637974745e-05, "loss": 0.0114, "step": 55430 }, { "grad_norm": 0.2028123140335083, "learning_rate": 2.36305490415076e-05, "loss": 0.0181, "step": 55440 }, { "grad_norm": 0.40026143193244934, "learning_rate": 2.3612990950858392e-05, "loss": 0.0139, "step": 55450 }, { "grad_norm": 0.2048855870962143, "learning_rate": 2.359543736902735e-05, "loss": 0.0159, "step": 55460 }, { "grad_norm": 0.16953083872795105, "learning_rate": 2.3577888299013896e-05, "loss": 0.0173, "step": 55470 }, { "grad_norm": 0.3409965932369232, "learning_rate": 2.3560343743816683e-05, "loss": 0.0153, "step": 55480 }, { "grad_norm": 0.343161404132843, "learning_rate": 2.354280370643362e-05, "loss": 0.0191, "step": 55490 }, { "grad_norm": 0.3196071684360504, "learning_rate": 2.3525268189861777e-05, "loss": 0.0163, "step": 55500 }, { "grad_norm": 0.31241095066070557, "learning_rate": 2.3507737197097513e-05, "loss": 0.013, "step": 55510 }, { "grad_norm": 0.34332239627838135, "learning_rate": 2.34902107311364e-05, "loss": 0.0172, "step": 55520 }, { "grad_norm": 0.20624244213104248, "learning_rate": 2.3472688794973225e-05, "loss": 0.017, "step": 55530 }, { "grad_norm": 0.3239201605319977, "learning_rate": 2.3455171391602016e-05, "loss": 0.0196, "step": 55540 }, { "grad_norm": 0.466448575258255, "learning_rate": 2.343765852401601e-05, "loss": 0.0157, "step": 55550 }, { "grad_norm": 0.18275868892669678, "learning_rate": 2.342015019520768e-05, "loss": 0.0185, "step": 55560 }, { "grad_norm": 0.2043193131685257, "learning_rate": 2.3402646408168742e-05, "loss": 0.0132, "step": 55570 }, { "grad_norm": 0.2826470732688904, "learning_rate": 2.3385147165890074e-05, "loss": 0.0192, "step": 55580 }, { "grad_norm": 0.266341894865036, "learning_rate": 2.336765247136184e-05, "loss": 0.0185, "step": 55590 }, { "grad_norm": 0.41428735852241516, "learning_rate": 2.33501623275734e-05, "loss": 0.0222, "step": 55600 }, { "grad_norm": 0.29477742314338684, "learning_rate": 2.333267673751334e-05, "loss": 0.0137, "step": 55610 }, { "grad_norm": 0.19788023829460144, "learning_rate": 2.33151957041695e-05, "loss": 0.0128, "step": 55620 }, { "grad_norm": 0.2500627040863037, "learning_rate": 2.329771923052884e-05, "loss": 0.0156, "step": 55630 }, { "grad_norm": 0.2443140745162964, "learning_rate": 2.3280247319577697e-05, "loss": 0.0135, "step": 55640 }, { "grad_norm": 0.4086655378341675, "learning_rate": 2.3262779974301473e-05, "loss": 0.0187, "step": 55650 }, { "grad_norm": 0.22950878739356995, "learning_rate": 2.3245317197684895e-05, "loss": 0.0154, "step": 55660 }, { "grad_norm": 0.3011705279350281, "learning_rate": 2.322785899271188e-05, "loss": 0.0129, "step": 55670 }, { "grad_norm": 0.29750627279281616, "learning_rate": 2.3210405362365507e-05, "loss": 0.0251, "step": 55680 }, { "grad_norm": 0.24051153659820557, "learning_rate": 2.31929563096282e-05, "loss": 0.0219, "step": 55690 }, { "grad_norm": 0.1619548350572586, "learning_rate": 2.317551183748146e-05, "loss": 0.0177, "step": 55700 }, { "grad_norm": 0.2046002894639969, "learning_rate": 2.3158071948906103e-05, "loss": 0.0154, "step": 55710 }, { "grad_norm": 0.19839592278003693, "learning_rate": 2.314063664688212e-05, "loss": 0.0155, "step": 55720 }, { "grad_norm": 0.29817503690719604, "learning_rate": 2.3123205934388725e-05, "loss": 0.0141, "step": 55730 }, { "grad_norm": 0.22890231013298035, "learning_rate": 2.310577981440436e-05, "loss": 0.0235, "step": 55740 }, { "grad_norm": 0.25336921215057373, "learning_rate": 2.3088358289906665e-05, "loss": 0.015, "step": 55750 }, { "grad_norm": 0.3127240538597107, "learning_rate": 2.307094136387252e-05, "loss": 0.016, "step": 55760 }, { "grad_norm": 0.22484657168388367, "learning_rate": 2.305352903927796e-05, "loss": 0.02, "step": 55770 }, { "grad_norm": 0.168714240193367, "learning_rate": 2.303612131909831e-05, "loss": 0.014, "step": 55780 }, { "grad_norm": 0.28308993577957153, "learning_rate": 2.3018718206308054e-05, "loss": 0.02, "step": 55790 }, { "grad_norm": 0.38735076785087585, "learning_rate": 2.3001319703880925e-05, "loss": 0.0219, "step": 55800 }, { "grad_norm": 0.23588158190250397, "learning_rate": 2.2983925814789835e-05, "loss": 0.0167, "step": 55810 }, { "grad_norm": 0.29197999835014343, "learning_rate": 2.2966536542006957e-05, "loss": 0.0171, "step": 55820 }, { "grad_norm": 0.1637984812259674, "learning_rate": 2.294915188850358e-05, "loss": 0.0161, "step": 55830 }, { "grad_norm": 0.2384343296289444, "learning_rate": 2.2931771857250333e-05, "loss": 0.0194, "step": 55840 }, { "grad_norm": 0.38134273886680603, "learning_rate": 2.2914396451216946e-05, "loss": 0.0171, "step": 55850 }, { "grad_norm": 0.3796137571334839, "learning_rate": 2.2897025673372412e-05, "loss": 0.015, "step": 55860 }, { "grad_norm": 0.24736590683460236, "learning_rate": 2.287965952668494e-05, "loss": 0.0192, "step": 55870 }, { "grad_norm": 0.1527697890996933, "learning_rate": 2.2862298014121873e-05, "loss": 0.0142, "step": 55880 }, { "grad_norm": 0.35238638520240784, "learning_rate": 2.28449411386499e-05, "loss": 0.0171, "step": 55890 }, { "grad_norm": 0.33013296127319336, "learning_rate": 2.282758890323477e-05, "loss": 0.017, "step": 55900 }, { "grad_norm": 0.4502801299095154, "learning_rate": 2.2810241310841528e-05, "loss": 0.015, "step": 55910 }, { "grad_norm": 0.20169095695018768, "learning_rate": 2.2792898364434412e-05, "loss": 0.018, "step": 55920 }, { "grad_norm": 0.16662171483039856, "learning_rate": 2.2775560066976843e-05, "loss": 0.0158, "step": 55930 }, { "grad_norm": 0.1592952013015747, "learning_rate": 2.275822642143147e-05, "loss": 0.0172, "step": 55940 }, { "grad_norm": 0.2109207957983017, "learning_rate": 2.274089743076014e-05, "loss": 0.0165, "step": 55950 }, { "grad_norm": 0.3073466718196869, "learning_rate": 2.2723573097923917e-05, "loss": 0.0153, "step": 55960 }, { "grad_norm": 0.28604868054389954, "learning_rate": 2.2706253425883024e-05, "loss": 0.0148, "step": 55970 }, { "grad_norm": 0.3581506311893463, "learning_rate": 2.2688938417596933e-05, "loss": 0.0148, "step": 55980 }, { "grad_norm": 0.15962284803390503, "learning_rate": 2.2671628076024305e-05, "loss": 0.0157, "step": 55990 }, { "grad_norm": 0.1758127510547638, "learning_rate": 2.2654322404123008e-05, "loss": 0.015, "step": 56000 }, { "grad_norm": 0.10947137326002121, "learning_rate": 2.2637021404850105e-05, "loss": 0.0115, "step": 56010 }, { "grad_norm": 0.36344674229621887, "learning_rate": 2.2619725081161885e-05, "loss": 0.0193, "step": 56020 }, { "grad_norm": 0.2575635612010956, "learning_rate": 2.260243343601376e-05, "loss": 0.0156, "step": 56030 }, { "grad_norm": 0.203993022441864, "learning_rate": 2.2585146472360473e-05, "loss": 0.0123, "step": 56040 }, { "grad_norm": 0.28304895758628845, "learning_rate": 2.2567864193155834e-05, "loss": 0.0156, "step": 56050 }, { "grad_norm": 0.38249242305755615, "learning_rate": 2.255058660135294e-05, "loss": 0.0193, "step": 56060 }, { "grad_norm": 0.4525153338909149, "learning_rate": 2.253331369990407e-05, "loss": 0.0235, "step": 56070 }, { "grad_norm": 0.29073411226272583, "learning_rate": 2.251604549176063e-05, "loss": 0.019, "step": 56080 }, { "grad_norm": 0.2339247167110443, "learning_rate": 2.249878197987337e-05, "loss": 0.017, "step": 56090 }, { "grad_norm": 0.2791726291179657, "learning_rate": 2.2481523167192087e-05, "loss": 0.0152, "step": 56100 }, { "grad_norm": 0.29911860823631287, "learning_rate": 2.2464269056665867e-05, "loss": 0.014, "step": 56110 }, { "grad_norm": 0.2741270363330841, "learning_rate": 2.2447019651242958e-05, "loss": 0.0247, "step": 56120 }, { "grad_norm": 0.27128222584724426, "learning_rate": 2.2429774953870818e-05, "loss": 0.0177, "step": 56130 }, { "grad_norm": 0.37060847878456116, "learning_rate": 2.241253496749611e-05, "loss": 0.0219, "step": 56140 }, { "grad_norm": 0.1710832417011261, "learning_rate": 2.2395299695064614e-05, "loss": 0.0135, "step": 56150 }, { "grad_norm": 0.2798948287963867, "learning_rate": 2.237806913952145e-05, "loss": 0.0167, "step": 56160 }, { "grad_norm": 0.2612070143222809, "learning_rate": 2.2360843303810798e-05, "loss": 0.0177, "step": 56170 }, { "grad_norm": 0.3618967831134796, "learning_rate": 2.2343622190876084e-05, "loss": 0.0155, "step": 56180 }, { "grad_norm": 0.268282949924469, "learning_rate": 2.2326405803659935e-05, "loss": 0.0182, "step": 56190 }, { "grad_norm": 0.22180062532424927, "learning_rate": 2.230919414510416e-05, "loss": 0.0173, "step": 56200 }, { "grad_norm": 0.39174696803092957, "learning_rate": 2.229198721814976e-05, "loss": 0.0175, "step": 56210 }, { "grad_norm": 0.2320479452610016, "learning_rate": 2.2274785025736948e-05, "loss": 0.014, "step": 56220 }, { "grad_norm": 0.12494217604398727, "learning_rate": 2.225758757080507e-05, "loss": 0.0175, "step": 56230 }, { "grad_norm": 0.4621908366680145, "learning_rate": 2.2240394856292723e-05, "loss": 0.019, "step": 56240 }, { "grad_norm": 0.42359161376953125, "learning_rate": 2.2223206885137664e-05, "loss": 0.0152, "step": 56250 }, { "grad_norm": 0.40936756134033203, "learning_rate": 2.2206023660276853e-05, "loss": 0.0172, "step": 56260 }, { "grad_norm": 0.4335991144180298, "learning_rate": 2.218884518464645e-05, "loss": 0.017, "step": 56270 }, { "grad_norm": 0.3100056052207947, "learning_rate": 2.2171671461181732e-05, "loss": 0.0175, "step": 56280 }, { "grad_norm": 0.2561594545841217, "learning_rate": 2.2154502492817292e-05, "loss": 0.0189, "step": 56290 }, { "grad_norm": 0.12699152529239655, "learning_rate": 2.2137338282486782e-05, "loss": 0.011, "step": 56300 }, { "grad_norm": 0.2544182538986206, "learning_rate": 2.2120178833123113e-05, "loss": 0.0198, "step": 56310 }, { "grad_norm": 0.33473271131515503, "learning_rate": 2.210302414765838e-05, "loss": 0.0158, "step": 56320 }, { "grad_norm": 0.20670798420906067, "learning_rate": 2.2085874229023808e-05, "loss": 0.0126, "step": 56330 }, { "grad_norm": 0.2981753945350647, "learning_rate": 2.2068729080149907e-05, "loss": 0.0142, "step": 56340 }, { "grad_norm": 0.440908819437027, "learning_rate": 2.205158870396625e-05, "loss": 0.0181, "step": 56350 }, { "grad_norm": 0.1471414715051651, "learning_rate": 2.2034453103401732e-05, "loss": 0.0164, "step": 56360 }, { "grad_norm": 0.15817321836948395, "learning_rate": 2.20173222813843e-05, "loss": 0.0122, "step": 56370 }, { "grad_norm": 0.20434273779392242, "learning_rate": 2.200019624084116e-05, "loss": 0.0184, "step": 56380 }, { "grad_norm": 0.2682957053184509, "learning_rate": 2.1983074984698687e-05, "loss": 0.0203, "step": 56390 }, { "grad_norm": 0.33960801362991333, "learning_rate": 2.1965958515882433e-05, "loss": 0.0175, "step": 56400 }, { "grad_norm": 0.17920146882534027, "learning_rate": 2.1948846837317162e-05, "loss": 0.0195, "step": 56410 }, { "grad_norm": 0.3096368908882141, "learning_rate": 2.1931739951926738e-05, "loss": 0.0132, "step": 56420 }, { "grad_norm": 0.23272277414798737, "learning_rate": 2.191463786263429e-05, "loss": 0.0093, "step": 56430 }, { "grad_norm": 0.19626756012439728, "learning_rate": 2.1897540572362095e-05, "loss": 0.0154, "step": 56440 }, { "grad_norm": 0.2181369662284851, "learning_rate": 2.1880448084031614e-05, "loss": 0.0135, "step": 56450 }, { "grad_norm": 0.2578970491886139, "learning_rate": 2.1863360400563482e-05, "loss": 0.0128, "step": 56460 }, { "grad_norm": 0.27000775933265686, "learning_rate": 2.184627752487754e-05, "loss": 0.0179, "step": 56470 }, { "grad_norm": 0.1814466416835785, "learning_rate": 2.1829199459892725e-05, "loss": 0.013, "step": 56480 }, { "grad_norm": 0.1880871057510376, "learning_rate": 2.1812126208527282e-05, "loss": 0.0156, "step": 56490 }, { "grad_norm": 0.33229002356529236, "learning_rate": 2.179505777369852e-05, "loss": 0.0195, "step": 56500 }, { "grad_norm": 0.26673856377601624, "learning_rate": 2.1777994158322974e-05, "loss": 0.0135, "step": 56510 }, { "grad_norm": 0.27907323837280273, "learning_rate": 2.176093536531637e-05, "loss": 0.0204, "step": 56520 }, { "grad_norm": 0.17896568775177002, "learning_rate": 2.1743881397593537e-05, "loss": 0.0134, "step": 56530 }, { "grad_norm": 0.34545081853866577, "learning_rate": 2.1726832258068595e-05, "loss": 0.0173, "step": 56540 }, { "grad_norm": 0.4130086898803711, "learning_rate": 2.170978794965472e-05, "loss": 0.0212, "step": 56550 }, { "grad_norm": 0.2735160291194916, "learning_rate": 2.169274847526438e-05, "loss": 0.0155, "step": 56560 }, { "grad_norm": 0.184539332985878, "learning_rate": 2.1675713837809103e-05, "loss": 0.0198, "step": 56570 }, { "grad_norm": 0.26367631554603577, "learning_rate": 2.1658684040199655e-05, "loss": 0.0166, "step": 56580 }, { "grad_norm": 0.2838573455810547, "learning_rate": 2.1641659085345974e-05, "loss": 0.0148, "step": 56590 }, { "grad_norm": 0.17021417617797852, "learning_rate": 2.1624638976157154e-05, "loss": 0.0217, "step": 56600 }, { "grad_norm": 0.591036319732666, "learning_rate": 2.1607623715541476e-05, "loss": 0.0233, "step": 56610 }, { "grad_norm": 0.2571595013141632, "learning_rate": 2.159061330640636e-05, "loss": 0.0163, "step": 56620 }, { "grad_norm": 0.1993381530046463, "learning_rate": 2.1573607751658425e-05, "loss": 0.0107, "step": 56630 }, { "grad_norm": 0.35592833161354065, "learning_rate": 2.155660705420347e-05, "loss": 0.0163, "step": 56640 }, { "grad_norm": 0.2517804503440857, "learning_rate": 2.153961121694644e-05, "loss": 0.017, "step": 56650 }, { "grad_norm": 0.2211763858795166, "learning_rate": 2.152262024279145e-05, "loss": 0.0186, "step": 56660 }, { "grad_norm": 0.42884182929992676, "learning_rate": 2.150563413464183e-05, "loss": 0.0183, "step": 56670 }, { "grad_norm": 0.2523695230484009, "learning_rate": 2.1488652895399974e-05, "loss": 0.0131, "step": 56680 }, { "grad_norm": 0.4363119304180145, "learning_rate": 2.1471676527967587e-05, "loss": 0.0138, "step": 56690 }, { "grad_norm": 0.34515514969825745, "learning_rate": 2.1454705035245414e-05, "loss": 0.0195, "step": 56700 }, { "grad_norm": 0.22708038985729218, "learning_rate": 2.143773842013343e-05, "loss": 0.0128, "step": 56710 }, { "grad_norm": 0.34960344433784485, "learning_rate": 2.1420776685530796e-05, "loss": 0.0193, "step": 56720 }, { "grad_norm": 0.26702243089675903, "learning_rate": 2.1403819834335742e-05, "loss": 0.0135, "step": 56730 }, { "grad_norm": 0.23538491129875183, "learning_rate": 2.1386867869445808e-05, "loss": 0.0191, "step": 56740 }, { "grad_norm": 0.2208949327468872, "learning_rate": 2.1369920793757548e-05, "loss": 0.0129, "step": 56750 }, { "grad_norm": 0.29021498560905457, "learning_rate": 2.1352978610166828e-05, "loss": 0.0137, "step": 56760 }, { "grad_norm": 0.22394150495529175, "learning_rate": 2.1336041321568546e-05, "loss": 0.0146, "step": 56770 }, { "grad_norm": 0.27228352427482605, "learning_rate": 2.131910893085684e-05, "loss": 0.0142, "step": 56780 }, { "grad_norm": 0.2153678983449936, "learning_rate": 2.1302181440925012e-05, "loss": 0.0188, "step": 56790 }, { "grad_norm": 0.17775271832942963, "learning_rate": 2.128525885466546e-05, "loss": 0.0158, "step": 56800 }, { "grad_norm": 0.20584845542907715, "learning_rate": 2.1268341174969847e-05, "loss": 0.0159, "step": 56810 }, { "grad_norm": 0.257895827293396, "learning_rate": 2.1251428404728894e-05, "loss": 0.0136, "step": 56820 }, { "grad_norm": 0.2249288409948349, "learning_rate": 2.123452054683256e-05, "loss": 0.0128, "step": 56830 }, { "grad_norm": 0.3746299743652344, "learning_rate": 2.1217617604169927e-05, "loss": 0.0178, "step": 56840 }, { "grad_norm": 0.41121628880500793, "learning_rate": 2.120071957962924e-05, "loss": 0.017, "step": 56850 }, { "grad_norm": 0.3136844336986542, "learning_rate": 2.1183826476097917e-05, "loss": 0.0242, "step": 56860 }, { "grad_norm": 0.2673318684101105, "learning_rate": 2.116693829646254e-05, "loss": 0.0191, "step": 56870 }, { "grad_norm": 0.3932344913482666, "learning_rate": 2.1150055043608806e-05, "loss": 0.0146, "step": 56880 }, { "grad_norm": 0.4659656882286072, "learning_rate": 2.113317672042162e-05, "loss": 0.0182, "step": 56890 }, { "grad_norm": 0.23933745920658112, "learning_rate": 2.1116303329785025e-05, "loss": 0.0147, "step": 56900 }, { "grad_norm": 0.37657806277275085, "learning_rate": 2.109943487458222e-05, "loss": 0.0202, "step": 56910 }, { "grad_norm": 0.25748854875564575, "learning_rate": 2.1082571357695574e-05, "loss": 0.0147, "step": 56920 }, { "grad_norm": 0.18436041474342346, "learning_rate": 2.1065712782006557e-05, "loss": 0.0142, "step": 56930 }, { "grad_norm": 0.15497808158397675, "learning_rate": 2.104885915039591e-05, "loss": 0.0153, "step": 56940 }, { "grad_norm": 0.2722814083099365, "learning_rate": 2.103201046574338e-05, "loss": 0.0122, "step": 56950 }, { "grad_norm": 0.35148563981056213, "learning_rate": 2.1015166730928022e-05, "loss": 0.0179, "step": 56960 }, { "grad_norm": 0.37574252486228943, "learning_rate": 2.0998327948827912e-05, "loss": 0.0208, "step": 56970 }, { "grad_norm": 0.37126463651657104, "learning_rate": 2.0981494122320363e-05, "loss": 0.0195, "step": 56980 }, { "grad_norm": 0.24390007555484772, "learning_rate": 2.0964665254281822e-05, "loss": 0.0187, "step": 56990 }, { "grad_norm": 0.29202771186828613, "learning_rate": 2.094784134758784e-05, "loss": 0.0181, "step": 57000 }, { "grad_norm": 0.2083154320716858, "learning_rate": 2.0931022405113226e-05, "loss": 0.0218, "step": 57010 }, { "grad_norm": 0.25627803802490234, "learning_rate": 2.091420842973183e-05, "loss": 0.0196, "step": 57020 }, { "grad_norm": 0.3894015848636627, "learning_rate": 2.0897399424316715e-05, "loss": 0.0118, "step": 57030 }, { "grad_norm": 0.3873237669467926, "learning_rate": 2.0880595391740078e-05, "loss": 0.0184, "step": 57040 }, { "grad_norm": 0.22202104330062866, "learning_rate": 2.0863796334873277e-05, "loss": 0.0168, "step": 57050 }, { "grad_norm": 0.21779322624206543, "learning_rate": 2.08470022565868e-05, "loss": 0.0112, "step": 57060 }, { "grad_norm": 0.21247747540473938, "learning_rate": 2.0830213159750317e-05, "loss": 0.0118, "step": 57070 }, { "grad_norm": 0.13468711078166962, "learning_rate": 2.0813429047232596e-05, "loss": 0.0156, "step": 57080 }, { "grad_norm": 0.15921826660633087, "learning_rate": 2.0796649921901594e-05, "loss": 0.0132, "step": 57090 }, { "grad_norm": 0.16476623713970184, "learning_rate": 2.077987578662441e-05, "loss": 0.0117, "step": 57100 }, { "grad_norm": 0.2663995921611786, "learning_rate": 2.0763106644267277e-05, "loss": 0.0184, "step": 57110 }, { "grad_norm": 0.29069557785987854, "learning_rate": 2.0746342497695607e-05, "loss": 0.0136, "step": 57120 }, { "grad_norm": 0.18499445915222168, "learning_rate": 2.0729583349773886e-05, "loss": 0.0117, "step": 57130 }, { "grad_norm": 0.3411465287208557, "learning_rate": 2.0712829203365853e-05, "loss": 0.0192, "step": 57140 }, { "grad_norm": 0.18508462607860565, "learning_rate": 2.0696080061334267e-05, "loss": 0.0174, "step": 57150 }, { "grad_norm": 0.27616921067237854, "learning_rate": 2.067933592654117e-05, "loss": 0.0184, "step": 57160 }, { "grad_norm": 0.314963698387146, "learning_rate": 2.066259680184763e-05, "loss": 0.0183, "step": 57170 }, { "grad_norm": 0.21156111359596252, "learning_rate": 2.0645862690113908e-05, "loss": 0.0127, "step": 57180 }, { "grad_norm": 0.27420681715011597, "learning_rate": 2.0629133594199436e-05, "loss": 0.0172, "step": 57190 }, { "grad_norm": 0.3130112886428833, "learning_rate": 2.0612409516962704e-05, "loss": 0.0138, "step": 57200 }, { "grad_norm": 0.31084150075912476, "learning_rate": 2.0595690461261467e-05, "loss": 0.0154, "step": 57210 }, { "grad_norm": 0.18313992023468018, "learning_rate": 2.0578976429952503e-05, "loss": 0.0107, "step": 57220 }, { "grad_norm": 0.36423906683921814, "learning_rate": 2.0562267425891802e-05, "loss": 0.0151, "step": 57230 }, { "grad_norm": 0.4053694009780884, "learning_rate": 2.0545563451934467e-05, "loss": 0.0163, "step": 57240 }, { "grad_norm": 0.3788023293018341, "learning_rate": 2.0528864510934764e-05, "loss": 0.0181, "step": 57250 }, { "grad_norm": 0.38283687829971313, "learning_rate": 2.0512170605746096e-05, "loss": 0.0231, "step": 57260 }, { "grad_norm": 0.18822914361953735, "learning_rate": 2.0495481739220963e-05, "loss": 0.0132, "step": 57270 }, { "grad_norm": 0.33738261461257935, "learning_rate": 2.0478797914211045e-05, "loss": 0.0181, "step": 57280 }, { "grad_norm": 0.25661784410476685, "learning_rate": 2.046211913356716e-05, "loss": 0.0185, "step": 57290 }, { "grad_norm": 0.28049755096435547, "learning_rate": 2.0445445400139247e-05, "loss": 0.0177, "step": 57300 }, { "grad_norm": 0.3978925049304962, "learning_rate": 2.0428776716776405e-05, "loss": 0.0162, "step": 57310 }, { "grad_norm": 0.37201985716819763, "learning_rate": 2.0412113086326856e-05, "loss": 0.0153, "step": 57320 }, { "grad_norm": 0.3241974711418152, "learning_rate": 2.0395454511637918e-05, "loss": 0.0126, "step": 57330 }, { "grad_norm": 0.4693126976490021, "learning_rate": 2.037880099555616e-05, "loss": 0.0159, "step": 57340 }, { "grad_norm": 0.22790879011154175, "learning_rate": 2.0362152540927144e-05, "loss": 0.0108, "step": 57350 }, { "grad_norm": 0.27705174684524536, "learning_rate": 2.0345509150595666e-05, "loss": 0.0212, "step": 57360 }, { "grad_norm": 0.3454652726650238, "learning_rate": 2.0328870827405617e-05, "loss": 0.0196, "step": 57370 }, { "grad_norm": 0.18060210347175598, "learning_rate": 2.0312237574200043e-05, "loss": 0.0108, "step": 57380 }, { "grad_norm": 0.34736624360084534, "learning_rate": 2.029560939382112e-05, "loss": 0.0181, "step": 57390 }, { "grad_norm": 0.2512587904930115, "learning_rate": 2.0278986289110097e-05, "loss": 0.0162, "step": 57400 }, { "grad_norm": 0.3260393738746643, "learning_rate": 2.0262368262907484e-05, "loss": 0.0171, "step": 57410 }, { "grad_norm": 0.3559572696685791, "learning_rate": 2.024575531805279e-05, "loss": 0.0206, "step": 57420 }, { "grad_norm": 0.2355005145072937, "learning_rate": 2.0229147457384735e-05, "loss": 0.0129, "step": 57430 }, { "grad_norm": 0.2114974856376648, "learning_rate": 2.0212544683741157e-05, "loss": 0.0224, "step": 57440 }, { "grad_norm": 0.2205156683921814, "learning_rate": 2.0195946999958976e-05, "loss": 0.0241, "step": 57450 }, { "grad_norm": 0.2372935265302658, "learning_rate": 2.017935440887434e-05, "loss": 0.0176, "step": 57460 }, { "grad_norm": 0.3377458155155182, "learning_rate": 2.0162766913322423e-05, "loss": 0.014, "step": 57470 }, { "grad_norm": 0.2837672829627991, "learning_rate": 2.0146184516137588e-05, "loss": 0.0189, "step": 57480 }, { "grad_norm": 0.25044742226600647, "learning_rate": 2.012960722015332e-05, "loss": 0.0163, "step": 57490 }, { "grad_norm": 0.33710387349128723, "learning_rate": 2.0113035028202214e-05, "loss": 0.0178, "step": 57500 }, { "grad_norm": 0.16990377008914948, "learning_rate": 2.009646794311602e-05, "loss": 0.0193, "step": 57510 }, { "grad_norm": 0.16930818557739258, "learning_rate": 2.007990596772559e-05, "loss": 0.0153, "step": 57520 }, { "grad_norm": 0.26738205552101135, "learning_rate": 2.0063349104860923e-05, "loss": 0.0174, "step": 57530 }, { "grad_norm": 0.1793033331632614, "learning_rate": 2.0046797357351116e-05, "loss": 0.0185, "step": 57540 }, { "grad_norm": 0.24207888543605804, "learning_rate": 2.0030250728024412e-05, "loss": 0.0126, "step": 57550 }, { "grad_norm": 0.20968958735466003, "learning_rate": 2.001370921970819e-05, "loss": 0.0113, "step": 57560 }, { "grad_norm": 0.29857930541038513, "learning_rate": 1.9997172835228932e-05, "loss": 0.0141, "step": 57570 }, { "grad_norm": 0.2820994555950165, "learning_rate": 1.9980641577412262e-05, "loss": 0.0133, "step": 57580 }, { "grad_norm": 0.4200180768966675, "learning_rate": 1.9964115449082925e-05, "loss": 0.0145, "step": 57590 }, { "grad_norm": 0.14974091947078705, "learning_rate": 1.9947594453064742e-05, "loss": 0.0171, "step": 57600 }, { "grad_norm": 0.27827176451683044, "learning_rate": 1.9931078592180774e-05, "loss": 0.0164, "step": 57610 }, { "grad_norm": 0.27729910612106323, "learning_rate": 1.9914567869253065e-05, "loss": 0.0162, "step": 57620 }, { "grad_norm": 0.33789119124412537, "learning_rate": 1.989806228710287e-05, "loss": 0.0215, "step": 57630 }, { "grad_norm": 0.19304338097572327, "learning_rate": 1.9881561848550555e-05, "loss": 0.0144, "step": 57640 }, { "grad_norm": 0.20495858788490295, "learning_rate": 1.9865066556415544e-05, "loss": 0.0193, "step": 57650 }, { "grad_norm": 0.2287057638168335, "learning_rate": 1.98485764135165e-05, "loss": 0.0153, "step": 57660 }, { "grad_norm": 0.1375814825296402, "learning_rate": 1.983209142267109e-05, "loss": 0.0152, "step": 57670 }, { "grad_norm": 0.3771778643131256, "learning_rate": 1.9815611586696165e-05, "loss": 0.0132, "step": 57680 }, { "grad_norm": 0.3109109699726105, "learning_rate": 1.9799136908407667e-05, "loss": 0.0141, "step": 57690 }, { "grad_norm": 0.1694633960723877, "learning_rate": 1.9782667390620678e-05, "loss": 0.0163, "step": 57700 }, { "grad_norm": 0.18903619050979614, "learning_rate": 1.976620303614939e-05, "loss": 0.0098, "step": 57710 }, { "grad_norm": 0.21729955077171326, "learning_rate": 1.9749743847807108e-05, "loss": 0.0154, "step": 57720 }, { "grad_norm": 0.4217420816421509, "learning_rate": 1.9733289828406272e-05, "loss": 0.0152, "step": 57730 }, { "grad_norm": 0.32060232758522034, "learning_rate": 1.9716840980758382e-05, "loss": 0.0169, "step": 57740 }, { "grad_norm": 0.1213122233748436, "learning_rate": 1.9700397307674134e-05, "loss": 0.0121, "step": 57750 }, { "grad_norm": 0.13764020800590515, "learning_rate": 1.968395881196328e-05, "loss": 0.0201, "step": 57760 }, { "grad_norm": 0.38669314980506897, "learning_rate": 1.966752549643473e-05, "loss": 0.0153, "step": 57770 }, { "grad_norm": 0.2237788885831833, "learning_rate": 1.965109736389647e-05, "loss": 0.0168, "step": 57780 }, { "grad_norm": 0.2559733986854553, "learning_rate": 1.9634674417155645e-05, "loss": 0.0125, "step": 57790 }, { "grad_norm": 0.28789597749710083, "learning_rate": 1.9618256659018434e-05, "loss": 0.0292, "step": 57800 }, { "grad_norm": 0.39764219522476196, "learning_rate": 1.9601844092290257e-05, "loss": 0.0167, "step": 57810 }, { "grad_norm": 0.30184611678123474, "learning_rate": 1.9585436719775512e-05, "loss": 0.0117, "step": 57820 }, { "grad_norm": 0.12063786387443542, "learning_rate": 1.9569034544277793e-05, "loss": 0.0135, "step": 57830 }, { "grad_norm": 0.16567674279212952, "learning_rate": 1.9552637568599798e-05, "loss": 0.0154, "step": 57840 }, { "grad_norm": 0.21807555854320526, "learning_rate": 1.953624579554327e-05, "loss": 0.0156, "step": 57850 }, { "grad_norm": 0.4403323531150818, "learning_rate": 1.951985922790918e-05, "loss": 0.0237, "step": 57860 }, { "grad_norm": 0.3738635182380676, "learning_rate": 1.9503477868497505e-05, "loss": 0.0134, "step": 57870 }, { "grad_norm": 0.3881153166294098, "learning_rate": 1.9487101720107375e-05, "loss": 0.0179, "step": 57880 }, { "grad_norm": 0.25224438309669495, "learning_rate": 1.9470730785537032e-05, "loss": 0.0131, "step": 57890 }, { "grad_norm": 0.35573506355285645, "learning_rate": 1.9454365067583823e-05, "loss": 0.0139, "step": 57900 }, { "grad_norm": 0.2929905354976654, "learning_rate": 1.9438004569044215e-05, "loss": 0.0141, "step": 57910 }, { "grad_norm": 0.21055011451244354, "learning_rate": 1.9421649292713724e-05, "loss": 0.0326, "step": 57920 }, { "grad_norm": 0.35557687282562256, "learning_rate": 1.9405299241387076e-05, "loss": 0.0163, "step": 57930 }, { "grad_norm": 0.20656920969486237, "learning_rate": 1.9388954417858007e-05, "loss": 0.0136, "step": 57940 }, { "grad_norm": 0.414722740650177, "learning_rate": 1.9372614824919417e-05, "loss": 0.0124, "step": 57950 }, { "grad_norm": 0.2877233624458313, "learning_rate": 1.9356280465363284e-05, "loss": 0.0243, "step": 57960 }, { "grad_norm": 0.266015887260437, "learning_rate": 1.9339951341980723e-05, "loss": 0.0128, "step": 57970 }, { "grad_norm": 0.26138874888420105, "learning_rate": 1.9323627457561916e-05, "loss": 0.0158, "step": 57980 }, { "grad_norm": 0.1725093573331833, "learning_rate": 1.9307308814896198e-05, "loss": 0.0085, "step": 57990 }, { "grad_norm": 0.33402150869369507, "learning_rate": 1.9290995416771935e-05, "loss": 0.0131, "step": 58000 }, { "grad_norm": 0.2611917555332184, "learning_rate": 1.9274687265976665e-05, "loss": 0.0161, "step": 58010 }, { "grad_norm": 0.3655378818511963, "learning_rate": 1.9258384365297e-05, "loss": 0.0169, "step": 58020 }, { "grad_norm": 0.27357128262519836, "learning_rate": 1.924208671751866e-05, "loss": 0.0176, "step": 58030 }, { "grad_norm": 0.25782501697540283, "learning_rate": 1.9225794325426492e-05, "loss": 0.0221, "step": 58040 }, { "grad_norm": 0.21807526051998138, "learning_rate": 1.920950719180436e-05, "loss": 0.0122, "step": 58050 }, { "grad_norm": 0.2411300390958786, "learning_rate": 1.919322531943536e-05, "loss": 0.0126, "step": 58060 }, { "grad_norm": 0.211488738656044, "learning_rate": 1.917694871110157e-05, "loss": 0.0176, "step": 58070 }, { "grad_norm": 0.35604196786880493, "learning_rate": 1.9160677369584234e-05, "loss": 0.0161, "step": 58080 }, { "grad_norm": 0.30900585651397705, "learning_rate": 1.9144411297663694e-05, "loss": 0.0188, "step": 58090 }, { "grad_norm": 0.2083754539489746, "learning_rate": 1.9128150498119328e-05, "loss": 0.0185, "step": 58100 }, { "grad_norm": 0.17998106777668, "learning_rate": 1.9111894973729726e-05, "loss": 0.0105, "step": 58110 }, { "grad_norm": 0.41603800654411316, "learning_rate": 1.9095644727272454e-05, "loss": 0.0178, "step": 58120 }, { "grad_norm": 0.36379578709602356, "learning_rate": 1.907939976152429e-05, "loss": 0.0184, "step": 58130 }, { "grad_norm": 0.3283882439136505, "learning_rate": 1.906316007926101e-05, "loss": 0.017, "step": 58140 }, { "grad_norm": 0.22228437662124634, "learning_rate": 1.904692568325755e-05, "loss": 0.0146, "step": 58150 }, { "grad_norm": 0.2694843113422394, "learning_rate": 1.9030696576287925e-05, "loss": 0.016, "step": 58160 }, { "grad_norm": 0.22836095094680786, "learning_rate": 1.9014472761125242e-05, "loss": 0.0148, "step": 58170 }, { "grad_norm": 0.19288596510887146, "learning_rate": 1.899825424054172e-05, "loss": 0.0143, "step": 58180 }, { "grad_norm": 0.2793569266796112, "learning_rate": 1.898204101730863e-05, "loss": 0.0137, "step": 58190 }, { "grad_norm": 0.24511104822158813, "learning_rate": 1.8965833094196394e-05, "loss": 0.0249, "step": 58200 }, { "grad_norm": 0.28595685958862305, "learning_rate": 1.8949630473974495e-05, "loss": 0.0174, "step": 58210 }, { "grad_norm": 0.26437118649482727, "learning_rate": 1.8933433159411517e-05, "loss": 0.0173, "step": 58220 }, { "grad_norm": 0.18582691252231598, "learning_rate": 1.891724115327514e-05, "loss": 0.0143, "step": 58230 }, { "grad_norm": 0.3387039303779602, "learning_rate": 1.8901054458332152e-05, "loss": 0.0194, "step": 58240 }, { "grad_norm": 0.11478367447853088, "learning_rate": 1.8884873077348364e-05, "loss": 0.0176, "step": 58250 }, { "grad_norm": 0.298586368560791, "learning_rate": 1.88686970130888e-05, "loss": 0.0166, "step": 58260 }, { "grad_norm": 0.27563753724098206, "learning_rate": 1.8852526268317456e-05, "loss": 0.0184, "step": 58270 }, { "grad_norm": 0.1544363647699356, "learning_rate": 1.883636084579749e-05, "loss": 0.0184, "step": 58280 }, { "grad_norm": 0.32872509956359863, "learning_rate": 1.8820200748291145e-05, "loss": 0.0146, "step": 58290 }, { "grad_norm": 0.1729954183101654, "learning_rate": 1.8804045978559686e-05, "loss": 0.0118, "step": 58300 }, { "grad_norm": 0.22529800236225128, "learning_rate": 1.8787896539363594e-05, "loss": 0.0159, "step": 58310 }, { "grad_norm": 0.11555837839841843, "learning_rate": 1.8771752433462298e-05, "loss": 0.0105, "step": 58320 }, { "grad_norm": 0.32973042130470276, "learning_rate": 1.8755613663614447e-05, "loss": 0.0149, "step": 58330 }, { "grad_norm": 0.3010280728340149, "learning_rate": 1.873948023257767e-05, "loss": 0.0157, "step": 58340 }, { "grad_norm": 0.4852198660373688, "learning_rate": 1.8723352143108747e-05, "loss": 0.0157, "step": 58350 }, { "grad_norm": 0.29721492528915405, "learning_rate": 1.870722939796352e-05, "loss": 0.0172, "step": 58360 }, { "grad_norm": 0.6509523391723633, "learning_rate": 1.869111199989693e-05, "loss": 0.017, "step": 58370 }, { "grad_norm": 0.2927767336368561, "learning_rate": 1.8674999951663018e-05, "loss": 0.0167, "step": 58380 }, { "grad_norm": 0.10863889753818512, "learning_rate": 1.8658893256014857e-05, "loss": 0.0137, "step": 58390 }, { "grad_norm": 0.20155617594718933, "learning_rate": 1.8642791915704655e-05, "loss": 0.0157, "step": 58400 }, { "grad_norm": 0.2948988378047943, "learning_rate": 1.8626695933483697e-05, "loss": 0.0155, "step": 58410 }, { "grad_norm": 0.24724049866199493, "learning_rate": 1.8610605312102335e-05, "loss": 0.0148, "step": 58420 }, { "grad_norm": 0.27678021788597107, "learning_rate": 1.8594520054310032e-05, "loss": 0.0172, "step": 58430 }, { "grad_norm": 0.25852900743484497, "learning_rate": 1.8578440162855326e-05, "loss": 0.0127, "step": 58440 }, { "grad_norm": 0.2627197802066803, "learning_rate": 1.8562365640485774e-05, "loss": 0.0161, "step": 58450 }, { "grad_norm": 0.3094509243965149, "learning_rate": 1.854629648994815e-05, "loss": 0.016, "step": 58460 }, { "grad_norm": 0.3096361756324768, "learning_rate": 1.8530232713988183e-05, "loss": 0.0137, "step": 58470 }, { "grad_norm": 0.2702553868293762, "learning_rate": 1.851417431535074e-05, "loss": 0.0149, "step": 58480 }, { "grad_norm": 0.2849840223789215, "learning_rate": 1.8498121296779785e-05, "loss": 0.0196, "step": 58490 }, { "grad_norm": 0.21083584427833557, "learning_rate": 1.848207366101829e-05, "loss": 0.0206, "step": 58500 }, { "grad_norm": 0.23382467031478882, "learning_rate": 1.8466031410808422e-05, "loss": 0.0125, "step": 58510 }, { "grad_norm": 0.43759897351264954, "learning_rate": 1.844999454889129e-05, "loss": 0.0163, "step": 58520 }, { "grad_norm": 0.3416699171066284, "learning_rate": 1.8433963078007237e-05, "loss": 0.0182, "step": 58530 }, { "grad_norm": 0.3203721344470978, "learning_rate": 1.841793700089554e-05, "loss": 0.0175, "step": 58540 }, { "grad_norm": 0.2526480257511139, "learning_rate": 1.8401916320294644e-05, "loss": 0.0145, "step": 58550 }, { "grad_norm": 0.3230469226837158, "learning_rate": 1.838590103894205e-05, "loss": 0.0167, "step": 58560 }, { "grad_norm": 0.2862892150878906, "learning_rate": 1.8369891159574292e-05, "loss": 0.0156, "step": 58570 }, { "grad_norm": 0.3090766966342926, "learning_rate": 1.835388668492708e-05, "loss": 0.0161, "step": 58580 }, { "grad_norm": 0.13564440608024597, "learning_rate": 1.8337887617735095e-05, "loss": 0.0148, "step": 58590 }, { "grad_norm": 0.19772611558437347, "learning_rate": 1.8321893960732157e-05, "loss": 0.0176, "step": 58600 }, { "grad_norm": 0.3252507746219635, "learning_rate": 1.8305905716651138e-05, "loss": 0.0149, "step": 58610 }, { "grad_norm": 0.16233184933662415, "learning_rate": 1.8289922888224e-05, "loss": 0.0133, "step": 58620 }, { "grad_norm": 0.30598267912864685, "learning_rate": 1.8273945478181765e-05, "loss": 0.0177, "step": 58630 }, { "grad_norm": 0.36168521642684937, "learning_rate": 1.8257973489254558e-05, "loss": 0.0161, "step": 58640 }, { "grad_norm": 0.21611502766609192, "learning_rate": 1.824200692417152e-05, "loss": 0.0161, "step": 58650 }, { "grad_norm": 0.23469719290733337, "learning_rate": 1.8226045785660912e-05, "loss": 0.0135, "step": 58660 }, { "grad_norm": 0.36816999316215515, "learning_rate": 1.821009007645006e-05, "loss": 0.0116, "step": 58670 }, { "grad_norm": 0.25797975063323975, "learning_rate": 1.8194139799265357e-05, "loss": 0.0105, "step": 58680 }, { "grad_norm": 0.4313090443611145, "learning_rate": 1.8178194956832295e-05, "loss": 0.0169, "step": 58690 }, { "grad_norm": 0.30822131037712097, "learning_rate": 1.8162255551875346e-05, "loss": 0.0166, "step": 58700 }, { "grad_norm": 0.21241208910942078, "learning_rate": 1.8146321587118194e-05, "loss": 0.018, "step": 58710 }, { "grad_norm": 0.2463836967945099, "learning_rate": 1.8130393065283448e-05, "loss": 0.0087, "step": 58720 }, { "grad_norm": 0.28581753373146057, "learning_rate": 1.8114469989092925e-05, "loss": 0.0148, "step": 58730 }, { "grad_norm": 0.3349756896495819, "learning_rate": 1.8098552361267397e-05, "loss": 0.0115, "step": 58740 }, { "grad_norm": 0.2295534759759903, "learning_rate": 1.8082640184526763e-05, "loss": 0.0117, "step": 58750 }, { "grad_norm": 0.2187780737876892, "learning_rate": 1.806673346158999e-05, "loss": 0.0136, "step": 58760 }, { "grad_norm": 0.15031170845031738, "learning_rate": 1.8050832195175067e-05, "loss": 0.0172, "step": 58770 }, { "grad_norm": 0.2849227488040924, "learning_rate": 1.8034936387999136e-05, "loss": 0.0173, "step": 58780 }, { "grad_norm": 0.15713609755039215, "learning_rate": 1.8019046042778315e-05, "loss": 0.0166, "step": 58790 }, { "grad_norm": 0.46857231855392456, "learning_rate": 1.800316116222785e-05, "loss": 0.0135, "step": 58800 }, { "grad_norm": 0.24334968626499176, "learning_rate": 1.7987281749062018e-05, "loss": 0.0361, "step": 58810 }, { "grad_norm": 0.22697308659553528, "learning_rate": 1.7971407805994195e-05, "loss": 0.0202, "step": 58820 }, { "grad_norm": 0.14382342994213104, "learning_rate": 1.7955539335736787e-05, "loss": 0.0181, "step": 58830 }, { "grad_norm": 0.19715432822704315, "learning_rate": 1.7939676341001304e-05, "loss": 0.0142, "step": 58840 }, { "grad_norm": 0.2993110418319702, "learning_rate": 1.7923818824498275e-05, "loss": 0.0136, "step": 58850 }, { "grad_norm": 0.24942456185817719, "learning_rate": 1.7907966788937315e-05, "loss": 0.0237, "step": 58860 }, { "grad_norm": 0.23501569032669067, "learning_rate": 1.7892120237027116e-05, "loss": 0.019, "step": 58870 }, { "grad_norm": 0.27512481808662415, "learning_rate": 1.7876279171475413e-05, "loss": 0.0186, "step": 58880 }, { "grad_norm": 0.26202768087387085, "learning_rate": 1.7860443594989028e-05, "loss": 0.0167, "step": 58890 }, { "grad_norm": 0.2507786452770233, "learning_rate": 1.784461351027379e-05, "loss": 0.0134, "step": 58900 }, { "grad_norm": 0.17610816657543182, "learning_rate": 1.7828788920034677e-05, "loss": 0.0228, "step": 58910 }, { "grad_norm": 0.404180645942688, "learning_rate": 1.7812969826975623e-05, "loss": 0.0168, "step": 58920 }, { "grad_norm": 0.22841158509254456, "learning_rate": 1.7797156233799738e-05, "loss": 0.0198, "step": 58930 }, { "grad_norm": 0.2564494013786316, "learning_rate": 1.7781348143209094e-05, "loss": 0.0145, "step": 58940 }, { "grad_norm": 0.2362402230501175, "learning_rate": 1.7765545557904862e-05, "loss": 0.0253, "step": 58950 }, { "grad_norm": 0.2763616740703583, "learning_rate": 1.7749748480587302e-05, "loss": 0.0133, "step": 58960 }, { "grad_norm": 0.23199258744716644, "learning_rate": 1.773395691395564e-05, "loss": 0.0121, "step": 58970 }, { "grad_norm": 0.16809633374214172, "learning_rate": 1.7718170860708305e-05, "loss": 0.0141, "step": 58980 }, { "grad_norm": 0.18302340805530548, "learning_rate": 1.770239032354264e-05, "loss": 0.0148, "step": 58990 }, { "grad_norm": 0.37035873532295227, "learning_rate": 1.7686615305155124e-05, "loss": 0.0177, "step": 59000 }, { "grad_norm": 0.24092918634414673, "learning_rate": 1.767084580824128e-05, "loss": 0.0091, "step": 59010 }, { "grad_norm": 0.5444216132164001, "learning_rate": 1.765508183549569e-05, "loss": 0.0231, "step": 59020 }, { "grad_norm": 0.3043360114097595, "learning_rate": 1.763932338961199e-05, "loss": 0.0159, "step": 59030 }, { "grad_norm": 0.1853814721107483, "learning_rate": 1.762357047328284e-05, "loss": 0.015, "step": 59040 }, { "grad_norm": 0.3088063895702362, "learning_rate": 1.7607823089199997e-05, "loss": 0.0223, "step": 59050 }, { "grad_norm": 0.22808124125003815, "learning_rate": 1.7592081240054265e-05, "loss": 0.0146, "step": 59060 }, { "grad_norm": 0.2717845141887665, "learning_rate": 1.7576344928535488e-05, "loss": 0.0151, "step": 59070 }, { "grad_norm": 0.2729984521865845, "learning_rate": 1.7560614157332573e-05, "loss": 0.0138, "step": 59080 }, { "grad_norm": 0.1907777488231659, "learning_rate": 1.7544888929133495e-05, "loss": 0.0125, "step": 59090 }, { "grad_norm": 0.2515565752983093, "learning_rate": 1.752916924662522e-05, "loss": 0.0187, "step": 59100 }, { "grad_norm": 0.32705119252204895, "learning_rate": 1.751345511249387e-05, "loss": 0.0119, "step": 59110 }, { "grad_norm": 0.2971618175506592, "learning_rate": 1.749774652942452e-05, "loss": 0.0145, "step": 59120 }, { "grad_norm": 0.2957627475261688, "learning_rate": 1.748204350010135e-05, "loss": 0.0161, "step": 59130 }, { "grad_norm": 0.17604343593120575, "learning_rate": 1.7466346027207574e-05, "loss": 0.0128, "step": 59140 }, { "grad_norm": 0.22741928696632385, "learning_rate": 1.745065411342547e-05, "loss": 0.0125, "step": 59150 }, { "grad_norm": 0.2526678144931793, "learning_rate": 1.7434967761436366e-05, "loss": 0.0173, "step": 59160 }, { "grad_norm": 0.331406831741333, "learning_rate": 1.741928697392058e-05, "loss": 0.0138, "step": 59170 }, { "grad_norm": 0.4895401895046234, "learning_rate": 1.7403611753557597e-05, "loss": 0.0146, "step": 59180 }, { "grad_norm": 0.32499226927757263, "learning_rate": 1.738794210302584e-05, "loss": 0.0117, "step": 59190 }, { "grad_norm": 0.15785953402519226, "learning_rate": 1.7372278025002837e-05, "loss": 0.0116, "step": 59200 }, { "grad_norm": 0.28137028217315674, "learning_rate": 1.7356619522165164e-05, "loss": 0.0145, "step": 59210 }, { "grad_norm": 0.17959237098693848, "learning_rate": 1.7340966597188378e-05, "loss": 0.0121, "step": 59220 }, { "grad_norm": 0.2571336328983307, "learning_rate": 1.732531925274722e-05, "loss": 0.0121, "step": 59230 }, { "grad_norm": 0.1667030155658722, "learning_rate": 1.7309677491515318e-05, "loss": 0.0157, "step": 59240 }, { "grad_norm": 0.3232748210430145, "learning_rate": 1.7294041316165455e-05, "loss": 0.0157, "step": 59250 }, { "grad_norm": 0.2946443259716034, "learning_rate": 1.727841072936942e-05, "loss": 0.0175, "step": 59260 }, { "grad_norm": 0.26729121804237366, "learning_rate": 1.7262785733798058e-05, "loss": 0.0117, "step": 59270 }, { "grad_norm": 0.23244185745716095, "learning_rate": 1.7247166332121246e-05, "loss": 0.0127, "step": 59280 }, { "grad_norm": 0.2083052098751068, "learning_rate": 1.7231552527007933e-05, "loss": 0.0147, "step": 59290 }, { "grad_norm": 0.2840092182159424, "learning_rate": 1.721594432112606e-05, "loss": 0.0147, "step": 59300 }, { "grad_norm": 0.17580381035804749, "learning_rate": 1.7200341717142655e-05, "loss": 0.0208, "step": 59310 }, { "grad_norm": 0.38006067276000977, "learning_rate": 1.718474471772377e-05, "loss": 0.0167, "step": 59320 }, { "grad_norm": 0.17666101455688477, "learning_rate": 1.7169153325534528e-05, "loss": 0.0124, "step": 59330 }, { "grad_norm": 0.2608506977558136, "learning_rate": 1.7153567543239045e-05, "loss": 0.0131, "step": 59340 }, { "grad_norm": 0.20413608849048615, "learning_rate": 1.7137987373500525e-05, "loss": 0.0152, "step": 59350 }, { "grad_norm": 0.36001282930374146, "learning_rate": 1.7122412818981198e-05, "loss": 0.0172, "step": 59360 }, { "grad_norm": 0.23595240712165833, "learning_rate": 1.7106843882342283e-05, "loss": 0.0135, "step": 59370 }, { "grad_norm": 0.15361155569553375, "learning_rate": 1.709128056624415e-05, "loss": 0.0148, "step": 59380 }, { "grad_norm": 0.28098779916763306, "learning_rate": 1.70757228733461e-05, "loss": 0.0113, "step": 59390 }, { "grad_norm": 0.18520748615264893, "learning_rate": 1.706017080630653e-05, "loss": 0.0099, "step": 59400 }, { "grad_norm": 0.30177730321884155, "learning_rate": 1.7044624367782873e-05, "loss": 0.0187, "step": 59410 }, { "grad_norm": 0.39374712109565735, "learning_rate": 1.7029083560431553e-05, "loss": 0.0157, "step": 59420 }, { "grad_norm": 0.16770273447036743, "learning_rate": 1.701354838690813e-05, "loss": 0.0176, "step": 59430 }, { "grad_norm": 0.2222912311553955, "learning_rate": 1.6998018849867087e-05, "loss": 0.0121, "step": 59440 }, { "grad_norm": 0.1933736950159073, "learning_rate": 1.698249495196202e-05, "loss": 0.015, "step": 59450 }, { "grad_norm": 0.28932490944862366, "learning_rate": 1.6966976695845528e-05, "loss": 0.0148, "step": 59460 }, { "grad_norm": 0.24505065381526947, "learning_rate": 1.6951464084169268e-05, "loss": 0.0134, "step": 59470 }, { "grad_norm": 0.21365614235401154, "learning_rate": 1.6935957119583916e-05, "loss": 0.0168, "step": 59480 }, { "grad_norm": 0.23693907260894775, "learning_rate": 1.6920455804739205e-05, "loss": 0.0142, "step": 59490 }, { "grad_norm": 0.3153976798057556, "learning_rate": 1.6904960142283855e-05, "loss": 0.0178, "step": 59500 }, { "grad_norm": 0.20374387502670288, "learning_rate": 1.6889470134865666e-05, "loss": 0.0132, "step": 59510 }, { "grad_norm": 0.29803580045700073, "learning_rate": 1.687398578513145e-05, "loss": 0.0173, "step": 59520 }, { "grad_norm": 0.23161335289478302, "learning_rate": 1.6858507095727066e-05, "loss": 0.0167, "step": 59530 }, { "grad_norm": 0.23371495306491852, "learning_rate": 1.6843034069297403e-05, "loss": 0.0131, "step": 59540 }, { "grad_norm": 0.31802767515182495, "learning_rate": 1.682756670848637e-05, "loss": 0.0248, "step": 59550 }, { "grad_norm": 0.2027001827955246, "learning_rate": 1.6812105015936937e-05, "loss": 0.0106, "step": 59560 }, { "grad_norm": 0.28158044815063477, "learning_rate": 1.6796648994291027e-05, "loss": 0.0183, "step": 59570 }, { "grad_norm": 0.18904666602611542, "learning_rate": 1.678119864618973e-05, "loss": 0.0096, "step": 59580 }, { "grad_norm": 0.2921159565448761, "learning_rate": 1.6765753974273023e-05, "loss": 0.0112, "step": 59590 }, { "grad_norm": 0.18489767611026764, "learning_rate": 1.675031498118001e-05, "loss": 0.0107, "step": 59600 }, { "grad_norm": 0.1413508504629135, "learning_rate": 1.6734881669548804e-05, "loss": 0.0127, "step": 59610 }, { "grad_norm": 0.37534502148628235, "learning_rate": 1.671945404201648e-05, "loss": 0.0231, "step": 59620 }, { "grad_norm": 0.22096943855285645, "learning_rate": 1.670403210121927e-05, "loss": 0.0097, "step": 59630 }, { "grad_norm": 0.2167074829339981, "learning_rate": 1.6688615849792312e-05, "loss": 0.0156, "step": 59640 }, { "grad_norm": 0.47285524010658264, "learning_rate": 1.6673205290369832e-05, "loss": 0.0208, "step": 59650 }, { "grad_norm": 0.33955439925193787, "learning_rate": 1.6657800425585086e-05, "loss": 0.0166, "step": 59660 }, { "grad_norm": 0.2899187505245209, "learning_rate": 1.6642401258070327e-05, "loss": 0.0296, "step": 59670 }, { "grad_norm": 0.2980528473854065, "learning_rate": 1.6627007790456884e-05, "loss": 0.0222, "step": 59680 }, { "grad_norm": 0.22438877820968628, "learning_rate": 1.6611620025375037e-05, "loss": 0.0113, "step": 59690 }, { "grad_norm": 0.26301199197769165, "learning_rate": 1.6596237965454154e-05, "loss": 0.0172, "step": 59700 }, { "grad_norm": 0.3697497248649597, "learning_rate": 1.6580861613322607e-05, "loss": 0.0147, "step": 59710 }, { "grad_norm": 0.4024368226528168, "learning_rate": 1.6565490971607796e-05, "loss": 0.0123, "step": 59720 }, { "grad_norm": 0.4062724709510803, "learning_rate": 1.655012604293614e-05, "loss": 0.0142, "step": 59730 }, { "grad_norm": 0.19188953936100006, "learning_rate": 1.653476682993309e-05, "loss": 0.0203, "step": 59740 }, { "grad_norm": 0.19963175058364868, "learning_rate": 1.651941333522311e-05, "loss": 0.0132, "step": 59750 }, { "grad_norm": 0.27730637788772583, "learning_rate": 1.6504065561429715e-05, "loss": 0.0159, "step": 59760 }, { "grad_norm": 0.2334843873977661, "learning_rate": 1.6488723511175385e-05, "loss": 0.0169, "step": 59770 }, { "grad_norm": 0.5016362071037292, "learning_rate": 1.6473387187081668e-05, "loss": 0.014, "step": 59780 }, { "grad_norm": 0.12954024970531464, "learning_rate": 1.6458056591769123e-05, "loss": 0.0111, "step": 59790 }, { "grad_norm": 0.24961699545383453, "learning_rate": 1.6442731727857335e-05, "loss": 0.0158, "step": 59800 }, { "grad_norm": 0.19081822037696838, "learning_rate": 1.642741259796492e-05, "loss": 0.0119, "step": 59810 }, { "grad_norm": 0.33038148283958435, "learning_rate": 1.641209920470944e-05, "loss": 0.0123, "step": 59820 }, { "grad_norm": 0.44170254468917847, "learning_rate": 1.639679155070762e-05, "loss": 0.0159, "step": 59830 }, { "grad_norm": 0.31630390882492065, "learning_rate": 1.6381489638575048e-05, "loss": 0.0156, "step": 59840 }, { "grad_norm": 0.18034759163856506, "learning_rate": 1.636619347092643e-05, "loss": 0.0133, "step": 59850 }, { "grad_norm": 0.2533535361289978, "learning_rate": 1.6350903050375476e-05, "loss": 0.0139, "step": 59860 }, { "grad_norm": 0.3410188555717468, "learning_rate": 1.6335618379534856e-05, "loss": 0.0146, "step": 59870 }, { "grad_norm": 0.2788277566432953, "learning_rate": 1.6320339461016364e-05, "loss": 0.0124, "step": 59880 }, { "grad_norm": 0.1668732911348343, "learning_rate": 1.6305066297430687e-05, "loss": 0.014, "step": 59890 }, { "grad_norm": 0.2171006053686142, "learning_rate": 1.6289798891387654e-05, "loss": 0.0176, "step": 59900 }, { "grad_norm": 0.23080269992351532, "learning_rate": 1.6274537245495995e-05, "loss": 0.0106, "step": 59910 }, { "grad_norm": 0.7745502591133118, "learning_rate": 1.6259281362363527e-05, "loss": 0.0135, "step": 59920 }, { "grad_norm": 0.294464111328125, "learning_rate": 1.6244031244597068e-05, "loss": 0.0227, "step": 59930 }, { "grad_norm": 0.22108471393585205, "learning_rate": 1.622878689480244e-05, "loss": 0.0136, "step": 59940 }, { "grad_norm": 0.4700162410736084, "learning_rate": 1.6213548315584498e-05, "loss": 0.0185, "step": 59950 }, { "grad_norm": 0.28024688363075256, "learning_rate": 1.6198315509547074e-05, "loss": 0.0154, "step": 59960 }, { "grad_norm": 1.117058277130127, "learning_rate": 1.6183088479293056e-05, "loss": 0.012, "step": 59970 }, { "grad_norm": 0.2870933711528778, "learning_rate": 1.6167867227424316e-05, "loss": 0.015, "step": 59980 }, { "grad_norm": 0.38424375653266907, "learning_rate": 1.615265175654176e-05, "loss": 0.0175, "step": 59990 }, { "grad_norm": 0.4676416516304016, "learning_rate": 1.613744206924529e-05, "loss": 0.0171, "step": 60000 }, { "grad_norm": 0.20409856736660004, "learning_rate": 1.6122238168133845e-05, "loss": 0.0142, "step": 60010 }, { "grad_norm": 0.21271342039108276, "learning_rate": 1.6107040055805305e-05, "loss": 0.0117, "step": 60020 }, { "grad_norm": 0.23400229215621948, "learning_rate": 1.6091847734856673e-05, "loss": 0.0182, "step": 60030 }, { "grad_norm": 0.29350265860557556, "learning_rate": 1.6076661207883865e-05, "loss": 0.0139, "step": 60040 }, { "grad_norm": 0.2543913722038269, "learning_rate": 1.6061480477481848e-05, "loss": 0.0154, "step": 60050 }, { "grad_norm": 0.28424668312072754, "learning_rate": 1.604630554624461e-05, "loss": 0.0114, "step": 60060 }, { "grad_norm": 0.4414110481739044, "learning_rate": 1.603113641676509e-05, "loss": 0.015, "step": 60070 }, { "grad_norm": 0.3169645071029663, "learning_rate": 1.6015973091635338e-05, "loss": 0.015, "step": 60080 }, { "grad_norm": 0.2966826558113098, "learning_rate": 1.6000815573446288e-05, "loss": 0.0133, "step": 60090 }, { "grad_norm": 0.16649672389030457, "learning_rate": 1.5985663864788002e-05, "loss": 0.0231, "step": 60100 }, { "grad_norm": 0.1724614053964615, "learning_rate": 1.597051796824946e-05, "loss": 0.0141, "step": 60110 }, { "grad_norm": 0.23028862476348877, "learning_rate": 1.5955377886418682e-05, "loss": 0.0183, "step": 60120 }, { "grad_norm": 0.41639453172683716, "learning_rate": 1.5940243621882704e-05, "loss": 0.0138, "step": 60130 }, { "grad_norm": 0.34114938974380493, "learning_rate": 1.5925115177227555e-05, "loss": 0.0188, "step": 60140 }, { "grad_norm": 0.355057030916214, "learning_rate": 1.5909992555038288e-05, "loss": 0.0157, "step": 60150 }, { "grad_norm": 1.1151889562606812, "learning_rate": 1.5894875757898912e-05, "loss": 0.0233, "step": 60160 }, { "grad_norm": 0.28443050384521484, "learning_rate": 1.58797647883925e-05, "loss": 0.0156, "step": 60170 }, { "grad_norm": 0.18589383363723755, "learning_rate": 1.586465964910109e-05, "loss": 0.0159, "step": 60180 }, { "grad_norm": 0.22097110748291016, "learning_rate": 1.5849560342605734e-05, "loss": 0.0144, "step": 60190 }, { "grad_norm": 0.18214064836502075, "learning_rate": 1.583446687148651e-05, "loss": 0.0111, "step": 60200 }, { "grad_norm": 0.14176228642463684, "learning_rate": 1.581937923832248e-05, "loss": 0.0149, "step": 60210 }, { "grad_norm": 0.22163541615009308, "learning_rate": 1.5804297445691663e-05, "loss": 0.0154, "step": 60220 }, { "grad_norm": 0.3136552572250366, "learning_rate": 1.578922149617119e-05, "loss": 0.0191, "step": 60230 }, { "grad_norm": 0.2382914125919342, "learning_rate": 1.5774151392337084e-05, "loss": 0.0149, "step": 60240 }, { "grad_norm": 0.24521870911121368, "learning_rate": 1.575908713676442e-05, "loss": 0.0139, "step": 60250 }, { "grad_norm": 0.4369446337223053, "learning_rate": 1.5744028732027293e-05, "loss": 0.0206, "step": 60260 }, { "grad_norm": 0.3156004250049591, "learning_rate": 1.5728976180698723e-05, "loss": 0.0152, "step": 60270 }, { "grad_norm": 0.2598693370819092, "learning_rate": 1.5713929485350836e-05, "loss": 0.0177, "step": 60280 }, { "grad_norm": 0.2720591127872467, "learning_rate": 1.569888864855465e-05, "loss": 0.0145, "step": 60290 }, { "grad_norm": 0.23179548978805542, "learning_rate": 1.568385367288028e-05, "loss": 0.0112, "step": 60300 }, { "grad_norm": 0.2016095519065857, "learning_rate": 1.5668824560896755e-05, "loss": 0.0153, "step": 60310 }, { "grad_norm": 0.171343594789505, "learning_rate": 1.5653801315172156e-05, "loss": 0.014, "step": 60320 }, { "grad_norm": 0.24229580163955688, "learning_rate": 1.5638783938273554e-05, "loss": 0.0108, "step": 60330 }, { "grad_norm": 0.385562002658844, "learning_rate": 1.5623772432766966e-05, "loss": 0.0137, "step": 60340 }, { "grad_norm": 0.3993677496910095, "learning_rate": 1.5608766801217507e-05, "loss": 0.0155, "step": 60350 }, { "grad_norm": 0.32780760526657104, "learning_rate": 1.5593767046189184e-05, "loss": 0.0139, "step": 60360 }, { "grad_norm": 0.24728304147720337, "learning_rate": 1.5578773170245064e-05, "loss": 0.0115, "step": 60370 }, { "grad_norm": 0.3759482502937317, "learning_rate": 1.5563785175947182e-05, "loss": 0.0213, "step": 60380 }, { "grad_norm": 0.21401351690292358, "learning_rate": 1.5548803065856582e-05, "loss": 0.0117, "step": 60390 }, { "grad_norm": 0.29818353056907654, "learning_rate": 1.55338268425333e-05, "loss": 0.0155, "step": 60400 }, { "grad_norm": 0.3195315897464752, "learning_rate": 1.5518856508536373e-05, "loss": 0.0142, "step": 60410 }, { "grad_norm": 0.3443228006362915, "learning_rate": 1.5503892066423786e-05, "loss": 0.0161, "step": 60420 }, { "grad_norm": 0.44471094012260437, "learning_rate": 1.548893351875258e-05, "loss": 0.022, "step": 60430 }, { "grad_norm": 0.34521859884262085, "learning_rate": 1.547398086807875e-05, "loss": 0.0189, "step": 60440 }, { "grad_norm": 0.28961169719696045, "learning_rate": 1.5459034116957305e-05, "loss": 0.0221, "step": 60450 }, { "grad_norm": 0.17798233032226562, "learning_rate": 1.544409326794225e-05, "loss": 0.0201, "step": 60460 }, { "grad_norm": 0.1588914841413498, "learning_rate": 1.542915832358651e-05, "loss": 0.0171, "step": 60470 }, { "grad_norm": 0.219073086977005, "learning_rate": 1.541422928644214e-05, "loss": 0.0133, "step": 60480 }, { "grad_norm": 0.31826603412628174, "learning_rate": 1.5399306159060024e-05, "loss": 0.021, "step": 60490 }, { "grad_norm": 0.2739660441875458, "learning_rate": 1.5384388943990185e-05, "loss": 0.0148, "step": 60500 }, { "grad_norm": 0.2943497896194458, "learning_rate": 1.5369477643781526e-05, "loss": 0.015, "step": 60510 }, { "grad_norm": 0.27949777245521545, "learning_rate": 1.5354572260981985e-05, "loss": 0.0148, "step": 60520 }, { "grad_norm": 0.25242942571640015, "learning_rate": 1.533967279813851e-05, "loss": 0.0101, "step": 60530 }, { "grad_norm": 0.28529268503189087, "learning_rate": 1.5324779257796956e-05, "loss": 0.0164, "step": 60540 }, { "grad_norm": 0.1455419510602951, "learning_rate": 1.5309891642502293e-05, "loss": 0.0161, "step": 60550 }, { "grad_norm": 0.2422688603401184, "learning_rate": 1.5295009954798357e-05, "loss": 0.0095, "step": 60560 }, { "grad_norm": 0.20492812991142273, "learning_rate": 1.5280134197228036e-05, "loss": 0.0141, "step": 60570 }, { "grad_norm": 0.176152303814888, "learning_rate": 1.526526437233319e-05, "loss": 0.0194, "step": 60580 }, { "grad_norm": 0.1665230095386505, "learning_rate": 1.525040048265466e-05, "loss": 0.0108, "step": 60590 }, { "grad_norm": 0.17563366889953613, "learning_rate": 1.5235542530732288e-05, "loss": 0.0121, "step": 60600 }, { "grad_norm": 0.14672519266605377, "learning_rate": 1.5220690519104901e-05, "loss": 0.0141, "step": 60610 }, { "grad_norm": 0.23652182519435883, "learning_rate": 1.5205844450310275e-05, "loss": 0.0154, "step": 60620 }, { "grad_norm": 0.18639947474002838, "learning_rate": 1.519100432688521e-05, "loss": 0.0146, "step": 60630 }, { "grad_norm": 0.44636183977127075, "learning_rate": 1.5176170151365476e-05, "loss": 0.0187, "step": 60640 }, { "grad_norm": 0.20239177346229553, "learning_rate": 1.5161341926285833e-05, "loss": 0.0105, "step": 60650 }, { "grad_norm": 0.16966211795806885, "learning_rate": 1.5146519654180025e-05, "loss": 0.0148, "step": 60660 }, { "grad_norm": 0.2451104372739792, "learning_rate": 1.5131703337580739e-05, "loss": 0.0172, "step": 60670 }, { "grad_norm": 0.21745124459266663, "learning_rate": 1.5116892979019731e-05, "loss": 0.0154, "step": 60680 }, { "grad_norm": 0.32177016139030457, "learning_rate": 1.5102088581027623e-05, "loss": 0.0137, "step": 60690 }, { "grad_norm": 0.23653262853622437, "learning_rate": 1.5087290146134154e-05, "loss": 0.0106, "step": 60700 }, { "grad_norm": 0.13808099925518036, "learning_rate": 1.5072497676867915e-05, "loss": 0.0123, "step": 60710 }, { "grad_norm": 0.3059331178665161, "learning_rate": 1.505771117575655e-05, "loss": 0.0165, "step": 60720 }, { "grad_norm": 0.2848622798919678, "learning_rate": 1.5042930645326691e-05, "loss": 0.0139, "step": 60730 }, { "grad_norm": 0.3609371483325958, "learning_rate": 1.5028156088103879e-05, "loss": 0.0153, "step": 60740 }, { "grad_norm": 0.4513581395149231, "learning_rate": 1.5013387506612735e-05, "loss": 0.0142, "step": 60750 }, { "grad_norm": 0.11912250518798828, "learning_rate": 1.4998624903376767e-05, "loss": 0.0146, "step": 60760 }, { "grad_norm": 0.46600809693336487, "learning_rate": 1.4983868280918517e-05, "loss": 0.0126, "step": 60770 }, { "grad_norm": 0.21390455961227417, "learning_rate": 1.4969117641759478e-05, "loss": 0.0174, "step": 60780 }, { "grad_norm": 0.18970297276973724, "learning_rate": 1.495437298842014e-05, "loss": 0.0166, "step": 60790 }, { "grad_norm": 0.26839321851730347, "learning_rate": 1.4939634323419976e-05, "loss": 0.0173, "step": 60800 }, { "grad_norm": 0.2493700385093689, "learning_rate": 1.4924901649277384e-05, "loss": 0.0165, "step": 60810 }, { "grad_norm": 0.3052263855934143, "learning_rate": 1.4910174968509793e-05, "loss": 0.0114, "step": 60820 }, { "grad_norm": 0.30154919624328613, "learning_rate": 1.4895454283633598e-05, "loss": 0.0163, "step": 60830 }, { "grad_norm": 0.28038719296455383, "learning_rate": 1.488073959716415e-05, "loss": 0.015, "step": 60840 }, { "grad_norm": 0.2465093582868576, "learning_rate": 1.4866030911615791e-05, "loss": 0.0159, "step": 60850 }, { "grad_norm": 0.19062626361846924, "learning_rate": 1.4851328229501849e-05, "loss": 0.0171, "step": 60860 }, { "grad_norm": 0.1865154653787613, "learning_rate": 1.4836631553334562e-05, "loss": 0.0114, "step": 60870 }, { "grad_norm": 0.3362656831741333, "learning_rate": 1.4821940885625251e-05, "loss": 0.015, "step": 60880 }, { "grad_norm": 0.19930370151996613, "learning_rate": 1.4807256228884109e-05, "loss": 0.0127, "step": 60890 }, { "grad_norm": 0.1679077446460724, "learning_rate": 1.4792577585620353e-05, "loss": 0.0141, "step": 60900 }, { "grad_norm": 0.347930908203125, "learning_rate": 1.4777904958342164e-05, "loss": 0.0211, "step": 60910 }, { "grad_norm": 0.19291169941425323, "learning_rate": 1.4763238349556691e-05, "loss": 0.0158, "step": 60920 }, { "grad_norm": 0.26348286867141724, "learning_rate": 1.4748577761770072e-05, "loss": 0.0179, "step": 60930 }, { "grad_norm": 0.16377343237400055, "learning_rate": 1.4733923197487354e-05, "loss": 0.0161, "step": 60940 }, { "grad_norm": 0.457859069108963, "learning_rate": 1.4719274659212662e-05, "loss": 0.0184, "step": 60950 }, { "grad_norm": 0.3096977770328522, "learning_rate": 1.4704632149448983e-05, "loss": 0.0169, "step": 60960 }, { "grad_norm": 0.3822857439517975, "learning_rate": 1.4689995670698343e-05, "loss": 0.0201, "step": 60970 }, { "grad_norm": 0.24020063877105713, "learning_rate": 1.4675365225461728e-05, "loss": 0.0211, "step": 60980 }, { "grad_norm": 0.6084727048873901, "learning_rate": 1.4660740816239032e-05, "loss": 0.0118, "step": 60990 }, { "grad_norm": 0.2672252655029297, "learning_rate": 1.464612244552923e-05, "loss": 0.0184, "step": 61000 }, { "grad_norm": 0.183198019862175, "learning_rate": 1.4631510115830161e-05, "loss": 0.0154, "step": 61010 }, { "grad_norm": 0.2551250159740448, "learning_rate": 1.4616903829638679e-05, "loss": 0.0101, "step": 61020 }, { "grad_norm": 0.12861958146095276, "learning_rate": 1.46023035894506e-05, "loss": 0.0153, "step": 61030 }, { "grad_norm": 0.3711068034172058, "learning_rate": 1.4587709397760713e-05, "loss": 0.0153, "step": 61040 }, { "grad_norm": 0.2064780592918396, "learning_rate": 1.4573121257062755e-05, "loss": 0.0166, "step": 61050 }, { "grad_norm": 0.3609282374382019, "learning_rate": 1.4558539169849472e-05, "loss": 0.0207, "step": 61060 }, { "grad_norm": 0.29089000821113586, "learning_rate": 1.4543963138612499e-05, "loss": 0.0137, "step": 61070 }, { "grad_norm": 0.24159716069698334, "learning_rate": 1.4529393165842498e-05, "loss": 0.0145, "step": 61080 }, { "grad_norm": 0.2729530930519104, "learning_rate": 1.4514829254029084e-05, "loss": 0.0169, "step": 61090 }, { "grad_norm": 0.1426822990179062, "learning_rate": 1.450027140566083e-05, "loss": 0.0112, "step": 61100 }, { "grad_norm": 0.5038332939147949, "learning_rate": 1.4485719623225268e-05, "loss": 0.014, "step": 61110 }, { "grad_norm": 0.22275076806545258, "learning_rate": 1.4471173909208912e-05, "loss": 0.0149, "step": 61120 }, { "grad_norm": 0.32894787192344666, "learning_rate": 1.4456634266097236e-05, "loss": 0.016, "step": 61130 }, { "grad_norm": 0.31658342480659485, "learning_rate": 1.444210069637461e-05, "loss": 0.0212, "step": 61140 }, { "grad_norm": 0.23115895688533783, "learning_rate": 1.4427573202524502e-05, "loss": 0.0091, "step": 61150 }, { "grad_norm": 0.15455931425094604, "learning_rate": 1.4413051787029208e-05, "loss": 0.0153, "step": 61160 }, { "grad_norm": 0.39394474029541016, "learning_rate": 1.439853645237006e-05, "loss": 0.0166, "step": 61170 }, { "grad_norm": 0.3110421299934387, "learning_rate": 1.4384027201027344e-05, "loss": 0.0147, "step": 61180 }, { "grad_norm": 0.17097163200378418, "learning_rate": 1.4369524035480253e-05, "loss": 0.0149, "step": 61190 }, { "grad_norm": 0.20357084274291992, "learning_rate": 1.4355026958207035e-05, "loss": 0.0128, "step": 61200 }, { "grad_norm": 0.2347077876329422, "learning_rate": 1.434053597168481e-05, "loss": 0.0142, "step": 61210 }, { "grad_norm": 0.23492102324962616, "learning_rate": 1.432605107838969e-05, "loss": 0.0161, "step": 61220 }, { "grad_norm": 0.3460600972175598, "learning_rate": 1.4311572280796759e-05, "loss": 0.02, "step": 61230 }, { "grad_norm": 0.2781580984592438, "learning_rate": 1.4297099581380047e-05, "loss": 0.0177, "step": 61240 }, { "grad_norm": 0.26373928785324097, "learning_rate": 1.4282632982612538e-05, "loss": 0.0158, "step": 61250 }, { "grad_norm": 0.2104441076517105, "learning_rate": 1.4268172486966197e-05, "loss": 0.0183, "step": 61260 }, { "grad_norm": 0.21713580191135406, "learning_rate": 1.4253718096911894e-05, "loss": 0.0148, "step": 61270 }, { "grad_norm": 0.3085079491138458, "learning_rate": 1.4239269814919504e-05, "loss": 0.0189, "step": 61280 }, { "grad_norm": 0.32819196581840515, "learning_rate": 1.4224827643457843e-05, "loss": 0.0251, "step": 61290 }, { "grad_norm": 0.28759458661079407, "learning_rate": 1.4210391584994686e-05, "loss": 0.0118, "step": 61300 }, { "grad_norm": 0.25206229090690613, "learning_rate": 1.4195961641996763e-05, "loss": 0.0165, "step": 61310 }, { "grad_norm": 0.322387158870697, "learning_rate": 1.4181537816929751e-05, "loss": 0.0154, "step": 61320 }, { "grad_norm": 0.35043612122535706, "learning_rate": 1.4167120112258303e-05, "loss": 0.0215, "step": 61330 }, { "grad_norm": 0.19464543461799622, "learning_rate": 1.4152708530445963e-05, "loss": 0.0111, "step": 61340 }, { "grad_norm": 0.23573102056980133, "learning_rate": 1.4138303073955344e-05, "loss": 0.0145, "step": 61350 }, { "grad_norm": 0.27863726019859314, "learning_rate": 1.4123903745247897e-05, "loss": 0.0135, "step": 61360 }, { "grad_norm": 0.17468379437923431, "learning_rate": 1.4109510546784082e-05, "loss": 0.0114, "step": 61370 }, { "grad_norm": 0.1507570594549179, "learning_rate": 1.4095123481023325e-05, "loss": 0.0174, "step": 61380 }, { "grad_norm": 0.25176507234573364, "learning_rate": 1.4080742550423936e-05, "loss": 0.0151, "step": 61390 }, { "grad_norm": 0.22325709462165833, "learning_rate": 1.4066367757443278e-05, "loss": 0.0134, "step": 61400 }, { "grad_norm": 0.37279796600341797, "learning_rate": 1.4051999104537572e-05, "loss": 0.0168, "step": 61410 }, { "grad_norm": 0.34229040145874023, "learning_rate": 1.4037636594162034e-05, "loss": 0.0139, "step": 61420 }, { "grad_norm": 0.2691514492034912, "learning_rate": 1.402328022877083e-05, "loss": 0.0137, "step": 61430 }, { "grad_norm": 0.1754385083913803, "learning_rate": 1.4008930010817073e-05, "loss": 0.017, "step": 61440 }, { "grad_norm": 0.2634623944759369, "learning_rate": 1.3994585942752831e-05, "loss": 0.0121, "step": 61450 }, { "grad_norm": 0.11832547932863235, "learning_rate": 1.398024802702909e-05, "loss": 0.0129, "step": 61460 }, { "grad_norm": 0.1746664196252823, "learning_rate": 1.3965916266095819e-05, "loss": 0.0157, "step": 61470 }, { "grad_norm": 0.35427534580230713, "learning_rate": 1.395159066240192e-05, "loss": 0.0156, "step": 61480 }, { "grad_norm": 0.24968843162059784, "learning_rate": 1.3937271218395248e-05, "loss": 0.0153, "step": 61490 }, { "grad_norm": 0.18372699618339539, "learning_rate": 1.3922957936522613e-05, "loss": 0.0154, "step": 61500 }, { "grad_norm": 0.5578176975250244, "learning_rate": 1.3908650819229758e-05, "loss": 0.0143, "step": 61510 }, { "grad_norm": 0.37100300192832947, "learning_rate": 1.389434986896137e-05, "loss": 0.0146, "step": 61520 }, { "grad_norm": 0.28597623109817505, "learning_rate": 1.388005508816112e-05, "loss": 0.0111, "step": 61530 }, { "grad_norm": 0.38896527886390686, "learning_rate": 1.3865766479271557e-05, "loss": 0.0168, "step": 61540 }, { "grad_norm": 0.2058282494544983, "learning_rate": 1.3851484044734225e-05, "loss": 0.02, "step": 61550 }, { "grad_norm": 0.2224210798740387, "learning_rate": 1.3837207786989608e-05, "loss": 0.0176, "step": 61560 }, { "grad_norm": 0.34909480810165405, "learning_rate": 1.382293770847713e-05, "loss": 0.0146, "step": 61570 }, { "grad_norm": 0.3968358635902405, "learning_rate": 1.3808673811635159e-05, "loss": 0.0138, "step": 61580 }, { "grad_norm": 0.24573098123073578, "learning_rate": 1.3794416098900975e-05, "loss": 0.0154, "step": 61590 }, { "grad_norm": 0.2412375658750534, "learning_rate": 1.3780164572710886e-05, "loss": 0.0175, "step": 61600 }, { "grad_norm": 0.2796213924884796, "learning_rate": 1.3765919235500035e-05, "loss": 0.0182, "step": 61610 }, { "grad_norm": 0.2486196756362915, "learning_rate": 1.375168008970259e-05, "loss": 0.0164, "step": 61620 }, { "grad_norm": 0.09483563154935837, "learning_rate": 1.3737447137751635e-05, "loss": 0.0161, "step": 61630 }, { "grad_norm": 0.35734468698501587, "learning_rate": 1.372322038207915e-05, "loss": 0.0149, "step": 61640 }, { "grad_norm": 0.2841867506504059, "learning_rate": 1.3708999825116159e-05, "loss": 0.0162, "step": 61650 }, { "grad_norm": 0.40616506338119507, "learning_rate": 1.3694785469292526e-05, "loss": 0.0196, "step": 61660 }, { "grad_norm": 0.29232174158096313, "learning_rate": 1.3680577317037101e-05, "loss": 0.0145, "step": 61670 }, { "grad_norm": 0.38627657294273376, "learning_rate": 1.3666375370777678e-05, "loss": 0.0151, "step": 61680 }, { "grad_norm": 0.3016960620880127, "learning_rate": 1.3652179632940981e-05, "loss": 0.0132, "step": 61690 }, { "grad_norm": 0.14797960221767426, "learning_rate": 1.3637990105952664e-05, "loss": 0.014, "step": 61700 }, { "grad_norm": 0.14852522313594818, "learning_rate": 1.3623806792237337e-05, "loss": 0.0148, "step": 61710 }, { "grad_norm": 0.3128149211406708, "learning_rate": 1.3609629694218551e-05, "loss": 0.017, "step": 61720 }, { "grad_norm": 0.1993507742881775, "learning_rate": 1.3595458814318762e-05, "loss": 0.0122, "step": 61730 }, { "grad_norm": 0.19158190488815308, "learning_rate": 1.3581294154959389e-05, "loss": 0.0163, "step": 61740 }, { "grad_norm": 0.32224830985069275, "learning_rate": 1.3567135718560792e-05, "loss": 0.013, "step": 61750 }, { "grad_norm": 0.24385160207748413, "learning_rate": 1.3552983507542261e-05, "loss": 0.0135, "step": 61760 }, { "grad_norm": 0.16425877809524536, "learning_rate": 1.3538837524322023e-05, "loss": 0.0171, "step": 61770 }, { "grad_norm": 0.1685757040977478, "learning_rate": 1.3524697771317251e-05, "loss": 0.0206, "step": 61780 }, { "grad_norm": 0.20062488317489624, "learning_rate": 1.3510564250943997e-05, "loss": 0.0144, "step": 61790 }, { "grad_norm": 0.20709143579006195, "learning_rate": 1.3496436965617353e-05, "loss": 0.0126, "step": 61800 }, { "grad_norm": 0.3217863142490387, "learning_rate": 1.3482315917751243e-05, "loss": 0.0172, "step": 61810 }, { "grad_norm": 0.2197093665599823, "learning_rate": 1.3468201109758582e-05, "loss": 0.0115, "step": 61820 }, { "grad_norm": 0.18651698529720306, "learning_rate": 1.3454092544051222e-05, "loss": 0.0173, "step": 61830 }, { "grad_norm": 0.2749011218547821, "learning_rate": 1.3439990223039878e-05, "loss": 0.0152, "step": 61840 }, { "grad_norm": 0.24351440370082855, "learning_rate": 1.3425894149134321e-05, "loss": 0.0137, "step": 61850 }, { "grad_norm": 0.2474825382232666, "learning_rate": 1.3411804324743132e-05, "loss": 0.0134, "step": 61860 }, { "grad_norm": 0.3057510554790497, "learning_rate": 1.3397720752273896e-05, "loss": 0.0173, "step": 61870 }, { "grad_norm": 0.2185123711824417, "learning_rate": 1.3383643434133108e-05, "loss": 0.0123, "step": 61880 }, { "grad_norm": 0.20316018164157867, "learning_rate": 1.3369572372726196e-05, "loss": 0.0139, "step": 61890 }, { "grad_norm": 0.20915676653385162, "learning_rate": 1.3355507570457525e-05, "loss": 0.0128, "step": 61900 }, { "grad_norm": 0.17391255497932434, "learning_rate": 1.334144902973038e-05, "loss": 0.013, "step": 61910 }, { "grad_norm": 0.11052007228136063, "learning_rate": 1.3327396752946997e-05, "loss": 0.0134, "step": 61920 }, { "grad_norm": 0.22424574196338654, "learning_rate": 1.3313350742508496e-05, "loss": 0.0123, "step": 61930 }, { "grad_norm": 0.2943948805332184, "learning_rate": 1.3299311000814973e-05, "loss": 0.0181, "step": 61940 }, { "grad_norm": 0.15500368177890778, "learning_rate": 1.3285277530265432e-05, "loss": 0.0161, "step": 61950 }, { "grad_norm": 0.36879095435142517, "learning_rate": 1.3271250333257813e-05, "loss": 0.0156, "step": 61960 }, { "grad_norm": 0.17238321900367737, "learning_rate": 1.3257229412188977e-05, "loss": 0.0108, "step": 61970 }, { "grad_norm": 0.3180568218231201, "learning_rate": 1.3243214769454726e-05, "loss": 0.0144, "step": 61980 }, { "grad_norm": 0.29102906584739685, "learning_rate": 1.3229206407449751e-05, "loss": 0.0116, "step": 61990 }, { "grad_norm": 0.21121656894683838, "learning_rate": 1.3215204328567742e-05, "loss": 0.0167, "step": 62000 }, { "grad_norm": 0.23531213402748108, "learning_rate": 1.3201208535201232e-05, "loss": 0.0128, "step": 62010 }, { "grad_norm": 0.18220734596252441, "learning_rate": 1.3187219029741737e-05, "loss": 0.0119, "step": 62020 }, { "grad_norm": 0.34104984998703003, "learning_rate": 1.3173235814579693e-05, "loss": 0.0123, "step": 62030 }, { "grad_norm": 0.2418333739042282, "learning_rate": 1.3159258892104398e-05, "loss": 0.0182, "step": 62040 }, { "grad_norm": 0.10972597450017929, "learning_rate": 1.3145288264704198e-05, "loss": 0.015, "step": 62050 }, { "grad_norm": 0.08711640536785126, "learning_rate": 1.3131323934766237e-05, "loss": 0.0107, "step": 62060 }, { "grad_norm": 0.18530778586864471, "learning_rate": 1.311736590467666e-05, "loss": 0.0147, "step": 62070 }, { "grad_norm": 0.3967234492301941, "learning_rate": 1.3103414176820505e-05, "loss": 0.0128, "step": 62080 }, { "grad_norm": 0.2558850049972534, "learning_rate": 1.308946875358174e-05, "loss": 0.0222, "step": 62090 }, { "grad_norm": 0.2258402556180954, "learning_rate": 1.3075529637343276e-05, "loss": 0.0223, "step": 62100 }, { "grad_norm": 0.27996188402175903, "learning_rate": 1.3061596830486883e-05, "loss": 0.0149, "step": 62110 }, { "grad_norm": 0.29751303791999817, "learning_rate": 1.3047670335393353e-05, "loss": 0.0135, "step": 62120 }, { "grad_norm": 0.211758092045784, "learning_rate": 1.3033750154442298e-05, "loss": 0.0208, "step": 62130 }, { "grad_norm": 0.5690183043479919, "learning_rate": 1.3019836290012316e-05, "loss": 0.0132, "step": 62140 }, { "grad_norm": 0.27477025985717773, "learning_rate": 1.30059287444809e-05, "loss": 0.0168, "step": 62150 }, { "grad_norm": 0.2456795573234558, "learning_rate": 1.299202752022447e-05, "loss": 0.0212, "step": 62160 }, { "grad_norm": 0.20140470564365387, "learning_rate": 1.2978132619618371e-05, "loss": 0.0108, "step": 62170 }, { "grad_norm": 0.33614301681518555, "learning_rate": 1.2964244045036866e-05, "loss": 0.0198, "step": 62180 }, { "grad_norm": 0.31012043356895447, "learning_rate": 1.295036179885311e-05, "loss": 0.0147, "step": 62190 }, { "grad_norm": 0.22017335891723633, "learning_rate": 1.293648588343922e-05, "loss": 0.0166, "step": 62200 }, { "grad_norm": 0.2598244249820709, "learning_rate": 1.2922616301166196e-05, "loss": 0.0163, "step": 62210 }, { "grad_norm": 0.34143978357315063, "learning_rate": 1.2908753054403976e-05, "loss": 0.0148, "step": 62220 }, { "grad_norm": 0.2453438639640808, "learning_rate": 1.2894896145521429e-05, "loss": 0.0122, "step": 62230 }, { "grad_norm": 0.20851829648017883, "learning_rate": 1.2881045576886275e-05, "loss": 0.0137, "step": 62240 }, { "grad_norm": 0.23895145952701569, "learning_rate": 1.2867201350865254e-05, "loss": 0.0193, "step": 62250 }, { "grad_norm": 0.20864619314670563, "learning_rate": 1.2853363469823914e-05, "loss": 0.0171, "step": 62260 }, { "grad_norm": 0.32297685742378235, "learning_rate": 1.283953193612682e-05, "loss": 0.0122, "step": 62270 }, { "grad_norm": 0.24095967411994934, "learning_rate": 1.2825706752137372e-05, "loss": 0.0139, "step": 62280 }, { "grad_norm": 0.4199589788913727, "learning_rate": 1.2811887920217896e-05, "loss": 0.0161, "step": 62290 }, { "grad_norm": 0.20473991334438324, "learning_rate": 1.2798075442729707e-05, "loss": 0.0115, "step": 62300 }, { "grad_norm": 0.2350596934556961, "learning_rate": 1.2784269322032922e-05, "loss": 0.0134, "step": 62310 }, { "grad_norm": 0.17652460932731628, "learning_rate": 1.2770469560486686e-05, "loss": 0.0147, "step": 62320 }, { "grad_norm": 0.2515304386615753, "learning_rate": 1.2756676160448956e-05, "loss": 0.0149, "step": 62330 }, { "grad_norm": 0.22168265283107758, "learning_rate": 1.2742889124276663e-05, "loss": 0.011, "step": 62340 }, { "grad_norm": 0.20776335895061493, "learning_rate": 1.2729108454325639e-05, "loss": 0.0151, "step": 62350 }, { "grad_norm": 0.30078282952308655, "learning_rate": 1.2715334152950614e-05, "loss": 0.0168, "step": 62360 }, { "grad_norm": 0.17603027820587158, "learning_rate": 1.2701566222505246e-05, "loss": 0.0141, "step": 62370 }, { "grad_norm": 0.3248140811920166, "learning_rate": 1.2687804665342107e-05, "loss": 0.0161, "step": 62380 }, { "grad_norm": 0.3406267464160919, "learning_rate": 1.267404948381265e-05, "loss": 0.014, "step": 62390 }, { "grad_norm": 0.1588204950094223, "learning_rate": 1.2660300680267267e-05, "loss": 0.0106, "step": 62400 }, { "grad_norm": 0.2516253888607025, "learning_rate": 1.2646558257055257e-05, "loss": 0.0131, "step": 62410 }, { "grad_norm": 0.2440183162689209, "learning_rate": 1.263282221652482e-05, "loss": 0.0101, "step": 62420 }, { "grad_norm": 0.20037445425987244, "learning_rate": 1.2619092561023088e-05, "loss": 0.0151, "step": 62430 }, { "grad_norm": 0.34655025601387024, "learning_rate": 1.2605369292896036e-05, "loss": 0.0167, "step": 62440 }, { "grad_norm": 0.2835560142993927, "learning_rate": 1.2591652414488658e-05, "loss": 0.0126, "step": 62450 }, { "grad_norm": 0.21310047805309296, "learning_rate": 1.2577941928144732e-05, "loss": 0.0143, "step": 62460 }, { "grad_norm": 0.20983651280403137, "learning_rate": 1.256423783620706e-05, "loss": 0.0146, "step": 62470 }, { "grad_norm": 0.2951469421386719, "learning_rate": 1.2550540141017264e-05, "loss": 0.0145, "step": 62480 }, { "grad_norm": 0.26212382316589355, "learning_rate": 1.253684884491591e-05, "loss": 0.0116, "step": 62490 }, { "grad_norm": 0.30118778347969055, "learning_rate": 1.2523163950242483e-05, "loss": 0.0122, "step": 62500 }, { "grad_norm": 0.2653799057006836, "learning_rate": 1.2509485459335313e-05, "loss": 0.0133, "step": 62510 }, { "grad_norm": 0.21046529710292816, "learning_rate": 1.2495813374531739e-05, "loss": 0.0125, "step": 62520 }, { "grad_norm": 0.24142755568027496, "learning_rate": 1.2482147698167907e-05, "loss": 0.0108, "step": 62530 }, { "grad_norm": 0.17016519606113434, "learning_rate": 1.2468488432578911e-05, "loss": 0.0137, "step": 62540 }, { "grad_norm": 0.23968924582004547, "learning_rate": 1.2454835580098761e-05, "loss": 0.0152, "step": 62550 }, { "grad_norm": 0.29919958114624023, "learning_rate": 1.2441189143060338e-05, "loss": 0.0145, "step": 62560 }, { "grad_norm": 0.24323387444019318, "learning_rate": 1.242754912379548e-05, "loss": 0.0144, "step": 62570 }, { "grad_norm": 0.46848952770233154, "learning_rate": 1.2413915524634844e-05, "loss": 0.0255, "step": 62580 }, { "grad_norm": 0.21612277626991272, "learning_rate": 1.2400288347908073e-05, "loss": 0.0153, "step": 62590 }, { "grad_norm": 0.24825911223888397, "learning_rate": 1.2386667595943663e-05, "loss": 0.0128, "step": 62600 }, { "grad_norm": 0.3578881621360779, "learning_rate": 1.2373053271069035e-05, "loss": 0.011, "step": 62610 }, { "grad_norm": 0.3081014156341553, "learning_rate": 1.2359445375610501e-05, "loss": 0.0145, "step": 62620 }, { "grad_norm": 0.17477834224700928, "learning_rate": 1.2345843911893301e-05, "loss": 0.0113, "step": 62630 }, { "grad_norm": 0.24046114087104797, "learning_rate": 1.2332248882241498e-05, "loss": 0.0222, "step": 62640 }, { "grad_norm": 0.28802916407585144, "learning_rate": 1.2318660288978178e-05, "loss": 0.0148, "step": 62650 }, { "grad_norm": 0.22588573396205902, "learning_rate": 1.2305078134425213e-05, "loss": 0.013, "step": 62660 }, { "grad_norm": 0.1867745816707611, "learning_rate": 1.2291502420903434e-05, "loss": 0.0132, "step": 62670 }, { "grad_norm": 0.33104127645492554, "learning_rate": 1.2277933150732567e-05, "loss": 0.0179, "step": 62680 }, { "grad_norm": 0.11888522654771805, "learning_rate": 1.2264370326231216e-05, "loss": 0.0115, "step": 62690 }, { "grad_norm": 0.33152130246162415, "learning_rate": 1.2250813949716927e-05, "loss": 0.0158, "step": 62700 }, { "grad_norm": 0.17096315324306488, "learning_rate": 1.2237264023506063e-05, "loss": 0.0153, "step": 62710 }, { "grad_norm": 0.5097498893737793, "learning_rate": 1.2223720549913987e-05, "loss": 0.012, "step": 62720 }, { "grad_norm": 0.24279507994651794, "learning_rate": 1.221018353125487e-05, "loss": 0.0142, "step": 62730 }, { "grad_norm": 0.191340833902359, "learning_rate": 1.2196652969841837e-05, "loss": 0.0123, "step": 62740 }, { "grad_norm": 0.2719957232475281, "learning_rate": 1.2183128867986904e-05, "loss": 0.0201, "step": 62750 }, { "grad_norm": 0.2930094301700592, "learning_rate": 1.2169611228000927e-05, "loss": 0.0126, "step": 62760 }, { "grad_norm": 0.2212776094675064, "learning_rate": 1.2156100052193752e-05, "loss": 0.0161, "step": 62770 }, { "grad_norm": 0.17966507375240326, "learning_rate": 1.214259534287403e-05, "loss": 0.0124, "step": 62780 }, { "grad_norm": 0.15376612544059753, "learning_rate": 1.2129097102349363e-05, "loss": 0.011, "step": 62790 }, { "grad_norm": 0.6033276319503784, "learning_rate": 1.2115605332926227e-05, "loss": 0.014, "step": 62800 }, { "grad_norm": 0.31923848390579224, "learning_rate": 1.2102120036910003e-05, "loss": 0.0234, "step": 62810 }, { "grad_norm": 0.25671735405921936, "learning_rate": 1.2088641216604956e-05, "loss": 0.0117, "step": 62820 }, { "grad_norm": 0.31929054856300354, "learning_rate": 1.2075168874314264e-05, "loss": 0.0128, "step": 62830 }, { "grad_norm": 0.11247965693473816, "learning_rate": 1.2061703012339942e-05, "loss": 0.0112, "step": 62840 }, { "grad_norm": 0.27601876854896545, "learning_rate": 1.204824363298297e-05, "loss": 0.0178, "step": 62850 }, { "grad_norm": 0.3579182028770447, "learning_rate": 1.2034790738543173e-05, "loss": 0.0162, "step": 62860 }, { "grad_norm": 0.21279649436473846, "learning_rate": 1.2021344331319284e-05, "loss": 0.0138, "step": 62870 }, { "grad_norm": 0.2797740697860718, "learning_rate": 1.200790441360894e-05, "loss": 0.0195, "step": 62880 }, { "grad_norm": 0.22589460015296936, "learning_rate": 1.199447098770864e-05, "loss": 0.0121, "step": 62890 }, { "grad_norm": 0.2797069847583771, "learning_rate": 1.198104405591381e-05, "loss": 0.0143, "step": 62900 }, { "grad_norm": 0.23431290686130524, "learning_rate": 1.1967623620518697e-05, "loss": 0.0103, "step": 62910 }, { "grad_norm": 0.30902695655822754, "learning_rate": 1.1954209683816554e-05, "loss": 0.013, "step": 62920 }, { "grad_norm": 0.17561277747154236, "learning_rate": 1.1940802248099402e-05, "loss": 0.0125, "step": 62930 }, { "grad_norm": 0.19795283675193787, "learning_rate": 1.1927401315658232e-05, "loss": 0.016, "step": 62940 }, { "grad_norm": 0.2780073583126068, "learning_rate": 1.1914006888782898e-05, "loss": 0.0161, "step": 62950 }, { "grad_norm": 0.2651066184043884, "learning_rate": 1.1900618969762106e-05, "loss": 0.0142, "step": 62960 }, { "grad_norm": 0.33751270174980164, "learning_rate": 1.188723756088354e-05, "loss": 0.017, "step": 62970 }, { "grad_norm": 0.23433077335357666, "learning_rate": 1.1873862664433672e-05, "loss": 0.0155, "step": 62980 }, { "grad_norm": 0.24844004213809967, "learning_rate": 1.1860494282697927e-05, "loss": 0.0125, "step": 62990 }, { "grad_norm": 0.31871500611305237, "learning_rate": 1.1847132417960588e-05, "loss": 0.0142, "step": 63000 }, { "grad_norm": 0.13151292502880096, "learning_rate": 1.1833777072504832e-05, "loss": 0.0159, "step": 63010 }, { "grad_norm": 0.21470940113067627, "learning_rate": 1.1820428248612731e-05, "loss": 0.014, "step": 63020 }, { "grad_norm": 0.4208163022994995, "learning_rate": 1.1807085948565245e-05, "loss": 0.0101, "step": 63030 }, { "grad_norm": 0.16734579205513, "learning_rate": 1.1793750174642172e-05, "loss": 0.0133, "step": 63040 }, { "grad_norm": 0.12179341167211533, "learning_rate": 1.1780420929122254e-05, "loss": 0.0132, "step": 63050 }, { "grad_norm": 0.21655678749084473, "learning_rate": 1.176709821428309e-05, "loss": 0.0136, "step": 63060 }, { "grad_norm": 0.19125811755657196, "learning_rate": 1.1753782032401173e-05, "loss": 0.0177, "step": 63070 }, { "grad_norm": 0.21507024765014648, "learning_rate": 1.1740472385751866e-05, "loss": 0.0155, "step": 63080 }, { "grad_norm": 0.14824627339839935, "learning_rate": 1.172716927660943e-05, "loss": 0.0106, "step": 63090 }, { "grad_norm": 0.3850648105144501, "learning_rate": 1.1713872707247015e-05, "loss": 0.0129, "step": 63100 }, { "grad_norm": 0.14956268668174744, "learning_rate": 1.1700582679936595e-05, "loss": 0.0133, "step": 63110 }, { "grad_norm": 0.2971460819244385, "learning_rate": 1.1687299196949136e-05, "loss": 0.0201, "step": 63120 }, { "grad_norm": 0.24335327744483948, "learning_rate": 1.1674022260554374e-05, "loss": 0.0125, "step": 63130 }, { "grad_norm": 0.2968839108943939, "learning_rate": 1.1660751873020987e-05, "loss": 0.011, "step": 63140 }, { "grad_norm": 0.3729178011417389, "learning_rate": 1.1647488036616538e-05, "loss": 0.015, "step": 63150 }, { "grad_norm": 0.2540030777454376, "learning_rate": 1.1634230753607417e-05, "loss": 0.0165, "step": 63160 }, { "grad_norm": 0.3712010383605957, "learning_rate": 1.1620980026258982e-05, "loss": 0.0136, "step": 63170 }, { "grad_norm": 0.33587419986724854, "learning_rate": 1.1607735856835373e-05, "loss": 0.0147, "step": 63180 }, { "grad_norm": 0.152613565325737, "learning_rate": 1.1594498247599677e-05, "loss": 0.0105, "step": 63190 }, { "grad_norm": 0.1885610669851303, "learning_rate": 1.158126720081384e-05, "loss": 0.0186, "step": 63200 }, { "grad_norm": 0.3847769796848297, "learning_rate": 1.156804271873868e-05, "loss": 0.0133, "step": 63210 }, { "grad_norm": 0.2980916202068329, "learning_rate": 1.1554824803633924e-05, "loss": 0.0156, "step": 63220 }, { "grad_norm": 0.2554868459701538, "learning_rate": 1.154161345775811e-05, "loss": 0.0118, "step": 63230 }, { "grad_norm": 0.16476282477378845, "learning_rate": 1.1528408683368724e-05, "loss": 0.0148, "step": 63240 }, { "grad_norm": 0.379256933927536, "learning_rate": 1.1515210482722088e-05, "loss": 0.0148, "step": 63250 }, { "grad_norm": 0.18712642788887024, "learning_rate": 1.150201885807342e-05, "loss": 0.0185, "step": 63260 }, { "grad_norm": 0.18105973303318024, "learning_rate": 1.1488833811676807e-05, "loss": 0.0125, "step": 63270 }, { "grad_norm": 0.2933575212955475, "learning_rate": 1.1475655345785213e-05, "loss": 0.014, "step": 63280 }, { "grad_norm": 0.25190427899360657, "learning_rate": 1.1462483462650481e-05, "loss": 0.0127, "step": 63290 }, { "grad_norm": 0.24820660054683685, "learning_rate": 1.1449318164523331e-05, "loss": 0.018, "step": 63300 }, { "grad_norm": 0.20786726474761963, "learning_rate": 1.1436159453653334e-05, "loss": 0.0084, "step": 63310 }, { "grad_norm": 0.24634428322315216, "learning_rate": 1.1423007332288955e-05, "loss": 0.0134, "step": 63320 }, { "grad_norm": 0.3897397518157959, "learning_rate": 1.1409861802677546e-05, "loss": 0.0155, "step": 63330 }, { "grad_norm": 0.3064892292022705, "learning_rate": 1.1396722867065313e-05, "loss": 0.0133, "step": 63340 }, { "grad_norm": 0.19263476133346558, "learning_rate": 1.1383590527697352e-05, "loss": 0.0129, "step": 63350 }, { "grad_norm": 0.17211849987506866, "learning_rate": 1.137046478681758e-05, "loss": 0.0102, "step": 63360 }, { "grad_norm": 0.3454062044620514, "learning_rate": 1.1357345646668888e-05, "loss": 0.0257, "step": 63370 }, { "grad_norm": 0.3213123679161072, "learning_rate": 1.1344233109492924e-05, "loss": 0.0155, "step": 63380 }, { "grad_norm": 0.31101351976394653, "learning_rate": 1.1331127177530292e-05, "loss": 0.0141, "step": 63390 }, { "grad_norm": 0.18974560499191284, "learning_rate": 1.1318027853020441e-05, "loss": 0.0141, "step": 63400 }, { "grad_norm": 0.2837243676185608, "learning_rate": 1.1304935138201645e-05, "loss": 0.0133, "step": 63410 }, { "grad_norm": 0.2430209517478943, "learning_rate": 1.1291849035311153e-05, "loss": 0.0132, "step": 63420 }, { "grad_norm": 0.18221774697303772, "learning_rate": 1.1278769546584972e-05, "loss": 0.0097, "step": 63430 }, { "grad_norm": 0.24136723577976227, "learning_rate": 1.1265696674258052e-05, "loss": 0.0118, "step": 63440 }, { "grad_norm": 0.19300460815429688, "learning_rate": 1.1252630420564186e-05, "loss": 0.0109, "step": 63450 }, { "grad_norm": 0.333601176738739, "learning_rate": 1.1239570787736036e-05, "loss": 0.0155, "step": 63460 }, { "grad_norm": 0.26749545335769653, "learning_rate": 1.1226517778005135e-05, "loss": 0.0127, "step": 63470 }, { "grad_norm": 0.22590549290180206, "learning_rate": 1.1213471393601893e-05, "loss": 0.0132, "step": 63480 }, { "grad_norm": 0.3452325761318207, "learning_rate": 1.1200431636755587e-05, "loss": 0.0149, "step": 63490 }, { "grad_norm": 0.3098818063735962, "learning_rate": 1.1187398509694336e-05, "loss": 0.0161, "step": 63500 }, { "grad_norm": 0.3816915452480316, "learning_rate": 1.1174372014645146e-05, "loss": 0.0119, "step": 63510 }, { "grad_norm": 0.15401692688465118, "learning_rate": 1.1161352153833899e-05, "loss": 0.0239, "step": 63520 }, { "grad_norm": 0.2566716969013214, "learning_rate": 1.1148338929485325e-05, "loss": 0.0113, "step": 63530 }, { "grad_norm": 0.07311646640300751, "learning_rate": 1.113533234382304e-05, "loss": 0.0107, "step": 63540 }, { "grad_norm": 0.2545984089374542, "learning_rate": 1.1122332399069513e-05, "loss": 0.0171, "step": 63550 }, { "grad_norm": 0.2266727238893509, "learning_rate": 1.1109339097446047e-05, "loss": 0.0126, "step": 63560 }, { "grad_norm": 0.3083098232746124, "learning_rate": 1.1096352441172897e-05, "loss": 0.0231, "step": 63570 }, { "grad_norm": 0.21414247155189514, "learning_rate": 1.1083372432469086e-05, "loss": 0.01, "step": 63580 }, { "grad_norm": 0.31130918860435486, "learning_rate": 1.107039907355255e-05, "loss": 0.0168, "step": 63590 }, { "grad_norm": 0.3826984167098999, "learning_rate": 1.1057432366640103e-05, "loss": 0.012, "step": 63600 }, { "grad_norm": 0.3837428689002991, "learning_rate": 1.1044472313947352e-05, "loss": 0.0167, "step": 63610 }, { "grad_norm": 0.26363101601600647, "learning_rate": 1.1031518917688877e-05, "loss": 0.0123, "step": 63620 }, { "grad_norm": 0.17436373233795166, "learning_rate": 1.1018572180078007e-05, "loss": 0.0229, "step": 63630 }, { "grad_norm": 0.32107171416282654, "learning_rate": 1.1005632103327018e-05, "loss": 0.0151, "step": 63640 }, { "grad_norm": 0.47249969840049744, "learning_rate": 1.0992698689646996e-05, "loss": 0.0155, "step": 63650 }, { "grad_norm": 0.2281920462846756, "learning_rate": 1.0979771941247919e-05, "loss": 0.0163, "step": 63660 }, { "grad_norm": 0.33499661087989807, "learning_rate": 1.0966851860338611e-05, "loss": 0.0118, "step": 63670 }, { "grad_norm": 0.26513412594795227, "learning_rate": 1.0953938449126766e-05, "loss": 0.0156, "step": 63680 }, { "grad_norm": 0.48540782928466797, "learning_rate": 1.0941031709818933e-05, "loss": 0.0247, "step": 63690 }, { "grad_norm": 0.34540146589279175, "learning_rate": 1.0928131644620509e-05, "loss": 0.0236, "step": 63700 }, { "grad_norm": 0.29308608174324036, "learning_rate": 1.0915238255735766e-05, "loss": 0.0128, "step": 63710 }, { "grad_norm": 0.6682210564613342, "learning_rate": 1.0902351545367833e-05, "loss": 0.0141, "step": 63720 }, { "grad_norm": 0.27439069747924805, "learning_rate": 1.0889471515718702e-05, "loss": 0.0113, "step": 63730 }, { "grad_norm": 0.22870786488056183, "learning_rate": 1.087659816898921e-05, "loss": 0.0102, "step": 63740 }, { "grad_norm": 0.27108630537986755, "learning_rate": 1.0863731507379082e-05, "loss": 0.0145, "step": 63750 }, { "grad_norm": 0.41652706265449524, "learning_rate": 1.0850871533086827e-05, "loss": 0.0162, "step": 63760 }, { "grad_norm": 0.2986880838871002, "learning_rate": 1.0838018248309927e-05, "loss": 0.016, "step": 63770 }, { "grad_norm": 0.5806682109832764, "learning_rate": 1.0825171655244615e-05, "loss": 0.0207, "step": 63780 }, { "grad_norm": 0.21912845969200134, "learning_rate": 1.0812331756086025e-05, "loss": 0.0136, "step": 63790 }, { "grad_norm": 0.4086844325065613, "learning_rate": 1.079949855302817e-05, "loss": 0.014, "step": 63800 }, { "grad_norm": 0.2134394496679306, "learning_rate": 1.0786672048263852e-05, "loss": 0.0158, "step": 63810 }, { "grad_norm": 0.30438825488090515, "learning_rate": 1.0773852243984817e-05, "loss": 0.0143, "step": 63820 }, { "grad_norm": 0.24919553101062775, "learning_rate": 1.0761039142381586e-05, "loss": 0.0162, "step": 63830 }, { "grad_norm": 0.3008529841899872, "learning_rate": 1.0748232745643577e-05, "loss": 0.0166, "step": 63840 }, { "grad_norm": 0.17515820264816284, "learning_rate": 1.0735433055959055e-05, "loss": 0.015, "step": 63850 }, { "grad_norm": 0.45882946252822876, "learning_rate": 1.0722640075515133e-05, "loss": 0.0139, "step": 63860 }, { "grad_norm": 0.18152618408203125, "learning_rate": 1.0709853806497795e-05, "loss": 0.0133, "step": 63870 }, { "grad_norm": 0.45006123185157776, "learning_rate": 1.0697074251091831e-05, "loss": 0.015, "step": 63880 }, { "grad_norm": 0.22593359649181366, "learning_rate": 1.0684301411480962e-05, "loss": 0.0117, "step": 63890 }, { "grad_norm": 0.1971648633480072, "learning_rate": 1.067153528984769e-05, "loss": 0.0109, "step": 63900 }, { "grad_norm": 0.27810826897621155, "learning_rate": 1.0658775888373395e-05, "loss": 0.0156, "step": 63910 }, { "grad_norm": 0.16298671066761017, "learning_rate": 1.0646023209238314e-05, "loss": 0.0147, "step": 63920 }, { "grad_norm": 0.445383220911026, "learning_rate": 1.0633277254621537e-05, "loss": 0.0147, "step": 63930 }, { "grad_norm": 0.22514158487319946, "learning_rate": 1.0620538026700994e-05, "loss": 0.0123, "step": 63940 }, { "grad_norm": 0.2027081847190857, "learning_rate": 1.0607805527653486e-05, "loss": 0.0141, "step": 63950 }, { "grad_norm": 0.33010783791542053, "learning_rate": 1.059507975965462e-05, "loss": 0.0162, "step": 63960 }, { "grad_norm": 0.11610428988933563, "learning_rate": 1.0582360724878898e-05, "loss": 0.0112, "step": 63970 }, { "grad_norm": 0.1920572966337204, "learning_rate": 1.0569648425499651e-05, "loss": 0.0144, "step": 63980 }, { "grad_norm": 0.18800175189971924, "learning_rate": 1.0556942863689063e-05, "loss": 0.0136, "step": 63990 }, { "grad_norm": 0.3196971118450165, "learning_rate": 1.054424404161819e-05, "loss": 0.015, "step": 64000 }, { "grad_norm": 0.17117731273174286, "learning_rate": 1.053155196145686e-05, "loss": 0.01, "step": 64010 }, { "grad_norm": 0.33196213841438293, "learning_rate": 1.0518866625373863e-05, "loss": 0.0132, "step": 64020 }, { "grad_norm": 0.14039665460586548, "learning_rate": 1.0506188035536735e-05, "loss": 0.0102, "step": 64030 }, { "grad_norm": 0.2208162397146225, "learning_rate": 1.0493516194111919e-05, "loss": 0.0172, "step": 64040 }, { "grad_norm": 0.20379211008548737, "learning_rate": 1.0480851103264688e-05, "loss": 0.0119, "step": 64050 }, { "grad_norm": 0.38704144954681396, "learning_rate": 1.046819276515913e-05, "loss": 0.0122, "step": 64060 }, { "grad_norm": 0.2640535235404968, "learning_rate": 1.0455541181958256e-05, "loss": 0.0136, "step": 64070 }, { "grad_norm": 0.3069503605365753, "learning_rate": 1.0442896355823822e-05, "loss": 0.0119, "step": 64080 }, { "grad_norm": 0.16843686997890472, "learning_rate": 1.0430258288916539e-05, "loss": 0.0136, "step": 64090 }, { "grad_norm": 0.2900950610637665, "learning_rate": 1.0417626983395868e-05, "loss": 0.014, "step": 64100 }, { "grad_norm": 0.17827871441841125, "learning_rate": 1.0405002441420165e-05, "loss": 0.0111, "step": 64110 }, { "grad_norm": 0.32149749994277954, "learning_rate": 1.039238466514662e-05, "loss": 0.0122, "step": 64120 }, { "grad_norm": 0.14584393799304962, "learning_rate": 1.0379773656731262e-05, "loss": 0.0135, "step": 64130 }, { "grad_norm": 0.27985385060310364, "learning_rate": 1.0367169418328986e-05, "loss": 0.0113, "step": 64140 }, { "grad_norm": 0.3662147521972656, "learning_rate": 1.0354571952093484e-05, "loss": 0.0158, "step": 64150 }, { "grad_norm": 0.23036719858646393, "learning_rate": 1.034198126017733e-05, "loss": 0.0123, "step": 64160 }, { "grad_norm": 0.14486022293567657, "learning_rate": 1.0329397344731928e-05, "loss": 0.0159, "step": 64170 }, { "grad_norm": 0.3508637249469757, "learning_rate": 1.0316820207907524e-05, "loss": 0.0149, "step": 64180 }, { "grad_norm": 0.1651630699634552, "learning_rate": 1.0304249851853215e-05, "loss": 0.0145, "step": 64190 }, { "grad_norm": 0.29368799924850464, "learning_rate": 1.0291686278716933e-05, "loss": 0.0119, "step": 64200 }, { "grad_norm": 0.23777464032173157, "learning_rate": 1.0279129490645418e-05, "loss": 0.0128, "step": 64210 }, { "grad_norm": 0.25925692915916443, "learning_rate": 1.0266579489784328e-05, "loss": 0.0128, "step": 64220 }, { "grad_norm": 0.3195416331291199, "learning_rate": 1.0254036278278084e-05, "loss": 0.0123, "step": 64230 }, { "grad_norm": 0.2881823182106018, "learning_rate": 1.0241499858269982e-05, "loss": 0.0135, "step": 64240 }, { "grad_norm": 0.2593044340610504, "learning_rate": 1.0228970231902169e-05, "loss": 0.0133, "step": 64250 }, { "grad_norm": 0.28372931480407715, "learning_rate": 1.0216447401315582e-05, "loss": 0.0163, "step": 64260 }, { "grad_norm": 0.25435739755630493, "learning_rate": 1.020393136865007e-05, "loss": 0.0098, "step": 64270 }, { "grad_norm": 0.31320059299468994, "learning_rate": 1.0191422136044242e-05, "loss": 0.0205, "step": 64280 }, { "grad_norm": 0.2733156383037567, "learning_rate": 1.017891970563563e-05, "loss": 0.0156, "step": 64290 }, { "grad_norm": 0.2209024429321289, "learning_rate": 1.0166424079560516e-05, "loss": 0.0237, "step": 64300 }, { "grad_norm": 0.3702254891395569, "learning_rate": 1.0153935259954078e-05, "loss": 0.016, "step": 64310 }, { "grad_norm": 0.4911327064037323, "learning_rate": 1.0141453248950311e-05, "loss": 0.0155, "step": 64320 }, { "grad_norm": 0.30509063601493835, "learning_rate": 1.0128978048682054e-05, "loss": 0.0162, "step": 64330 }, { "grad_norm": 0.3337680697441101, "learning_rate": 1.0116509661280982e-05, "loss": 0.0204, "step": 64340 }, { "grad_norm": 0.10096163302659988, "learning_rate": 1.0104048088877576e-05, "loss": 0.0104, "step": 64350 }, { "grad_norm": 0.2705370783805847, "learning_rate": 1.0091593333601201e-05, "loss": 0.0115, "step": 64360 }, { "grad_norm": 0.25937601923942566, "learning_rate": 1.0079145397580031e-05, "loss": 0.0137, "step": 64370 }, { "grad_norm": 0.3502487540245056, "learning_rate": 1.006670428294107e-05, "loss": 0.0187, "step": 64380 }, { "grad_norm": 0.21068820357322693, "learning_rate": 1.0054269991810166e-05, "loss": 0.0112, "step": 64390 }, { "grad_norm": 0.3617171347141266, "learning_rate": 1.0041842526312024e-05, "loss": 0.0168, "step": 64400 }, { "grad_norm": 0.12602464854717255, "learning_rate": 1.0029421888570101e-05, "loss": 0.0105, "step": 64410 }, { "grad_norm": 0.16250282526016235, "learning_rate": 1.0017008080706813e-05, "loss": 0.0222, "step": 64420 }, { "grad_norm": 0.10777066648006439, "learning_rate": 1.0004601104843287e-05, "loss": 0.0114, "step": 64430 }, { "grad_norm": 0.2631247937679291, "learning_rate": 9.992200963099562e-06, "loss": 0.0123, "step": 64440 }, { "grad_norm": 0.16426491737365723, "learning_rate": 9.979807657594486e-06, "loss": 0.0117, "step": 64450 }, { "grad_norm": 0.35209164023399353, "learning_rate": 9.967421190445703e-06, "loss": 0.0159, "step": 64460 }, { "grad_norm": 0.2416413128376007, "learning_rate": 9.955041563769769e-06, "loss": 0.0137, "step": 64470 }, { "grad_norm": 0.17396892607212067, "learning_rate": 9.942668779681974e-06, "loss": 0.0158, "step": 64480 }, { "grad_norm": 0.18913735449314117, "learning_rate": 9.930302840296541e-06, "loss": 0.0118, "step": 64490 }, { "grad_norm": 0.13904184103012085, "learning_rate": 9.917943747726426e-06, "loss": 0.0111, "step": 64500 }, { "grad_norm": 0.2938961982727051, "learning_rate": 9.905591504083484e-06, "loss": 0.0134, "step": 64510 }, { "grad_norm": 0.2574491500854492, "learning_rate": 9.893246111478382e-06, "loss": 0.0136, "step": 64520 }, { "grad_norm": 0.314303457736969, "learning_rate": 9.88090757202057e-06, "loss": 0.0108, "step": 64530 }, { "grad_norm": 0.33902406692504883, "learning_rate": 9.868575887818421e-06, "loss": 0.0157, "step": 64540 }, { "grad_norm": 0.19238315522670746, "learning_rate": 9.856251060979044e-06, "loss": 0.0122, "step": 64550 }, { "grad_norm": 0.2947361469268799, "learning_rate": 9.843933093608426e-06, "loss": 0.0169, "step": 64560 }, { "grad_norm": 0.3681906461715698, "learning_rate": 9.831621987811368e-06, "loss": 0.0167, "step": 64570 }, { "grad_norm": 0.21070095896720886, "learning_rate": 9.819317745691509e-06, "loss": 0.0129, "step": 64580 }, { "grad_norm": 0.36669084429740906, "learning_rate": 9.8070203693513e-06, "loss": 0.014, "step": 64590 }, { "grad_norm": 0.3309507668018341, "learning_rate": 9.794729860892048e-06, "loss": 0.0122, "step": 64600 }, { "grad_norm": 0.2785121500492096, "learning_rate": 9.782446222413827e-06, "loss": 0.0215, "step": 64610 }, { "grad_norm": 0.38274529576301575, "learning_rate": 9.770169456015598e-06, "loss": 0.018, "step": 64620 }, { "grad_norm": 0.15482939779758453, "learning_rate": 9.75789956379512e-06, "loss": 0.012, "step": 64630 }, { "grad_norm": 0.268989235162735, "learning_rate": 9.74563654784898e-06, "loss": 0.013, "step": 64640 }, { "grad_norm": 0.3024705946445465, "learning_rate": 9.733380410272596e-06, "loss": 0.0136, "step": 64650 }, { "grad_norm": 0.3315345346927643, "learning_rate": 9.721131153160207e-06, "loss": 0.0147, "step": 64660 }, { "grad_norm": 0.1641232967376709, "learning_rate": 9.708888778604886e-06, "loss": 0.0144, "step": 64670 }, { "grad_norm": 0.47483330965042114, "learning_rate": 9.696653288698488e-06, "loss": 0.0122, "step": 64680 }, { "grad_norm": 0.15294480323791504, "learning_rate": 9.684424685531763e-06, "loss": 0.0189, "step": 64690 }, { "grad_norm": 0.21289975941181183, "learning_rate": 9.672202971194216e-06, "loss": 0.0183, "step": 64700 }, { "grad_norm": 0.1576659083366394, "learning_rate": 9.659988147774213e-06, "loss": 0.0179, "step": 64710 }, { "grad_norm": 0.15907785296440125, "learning_rate": 9.647780217358942e-06, "loss": 0.0173, "step": 64720 }, { "grad_norm": 0.1982179582118988, "learning_rate": 9.635579182034376e-06, "loss": 0.0153, "step": 64730 }, { "grad_norm": 0.17046889662742615, "learning_rate": 9.623385043885386e-06, "loss": 0.0168, "step": 64740 }, { "grad_norm": 0.29381924867630005, "learning_rate": 9.61119780499557e-06, "loss": 0.0089, "step": 64750 }, { "grad_norm": 0.2912858724594116, "learning_rate": 9.599017467447418e-06, "loss": 0.0144, "step": 64760 }, { "grad_norm": 0.2517346441745758, "learning_rate": 9.586844033322206e-06, "loss": 0.0169, "step": 64770 }, { "grad_norm": 0.48680418729782104, "learning_rate": 9.574677504700052e-06, "loss": 0.0202, "step": 64780 }, { "grad_norm": 0.25473397970199585, "learning_rate": 9.562517883659877e-06, "loss": 0.0213, "step": 64790 }, { "grad_norm": 0.2573975622653961, "learning_rate": 9.55036517227944e-06, "loss": 0.0195, "step": 64800 }, { "grad_norm": 0.332968533039093, "learning_rate": 9.538219372635282e-06, "loss": 0.0133, "step": 64810 }, { "grad_norm": 0.21254996955394745, "learning_rate": 9.526080486802802e-06, "loss": 0.0127, "step": 64820 }, { "grad_norm": 0.2569349408149719, "learning_rate": 9.513948516856203e-06, "loss": 0.0163, "step": 64830 }, { "grad_norm": 0.15279537439346313, "learning_rate": 9.501823464868504e-06, "loss": 0.0113, "step": 64840 }, { "grad_norm": 0.2607681155204773, "learning_rate": 9.489705332911547e-06, "loss": 0.0115, "step": 64850 }, { "grad_norm": 0.317013144493103, "learning_rate": 9.477594123055994e-06, "loss": 0.016, "step": 64860 }, { "grad_norm": 0.1496279239654541, "learning_rate": 9.465489837371321e-06, "loss": 0.02, "step": 64870 }, { "grad_norm": 0.26048946380615234, "learning_rate": 9.453392477925794e-06, "loss": 0.0132, "step": 64880 }, { "grad_norm": 0.18340368568897247, "learning_rate": 9.441302046786566e-06, "loss": 0.0164, "step": 64890 }, { "grad_norm": 0.30786964297294617, "learning_rate": 9.429218546019519e-06, "loss": 0.0128, "step": 64900 }, { "grad_norm": 0.26937413215637207, "learning_rate": 9.41714197768941e-06, "loss": 0.0214, "step": 64910 }, { "grad_norm": 0.29111194610595703, "learning_rate": 9.405072343859805e-06, "loss": 0.0138, "step": 64920 }, { "grad_norm": 0.24950768053531647, "learning_rate": 9.393009646593043e-06, "loss": 0.014, "step": 64930 }, { "grad_norm": 0.24414585530757904, "learning_rate": 9.38095388795035e-06, "loss": 0.0151, "step": 64940 }, { "grad_norm": 0.34454914927482605, "learning_rate": 9.36890506999169e-06, "loss": 0.0139, "step": 64950 }, { "grad_norm": 0.2047845870256424, "learning_rate": 9.356863194775894e-06, "loss": 0.0158, "step": 64960 }, { "grad_norm": 0.31068864464759827, "learning_rate": 9.344828264360583e-06, "loss": 0.0166, "step": 64970 }, { "grad_norm": 0.1619359254837036, "learning_rate": 9.332800280802201e-06, "loss": 0.0174, "step": 64980 }, { "grad_norm": 0.11755365878343582, "learning_rate": 9.32077924615602e-06, "loss": 0.0138, "step": 64990 }, { "grad_norm": 0.24200594425201416, "learning_rate": 9.308765162476063e-06, "loss": 0.013, "step": 65000 }, { "grad_norm": 0.35302576422691345, "learning_rate": 9.296758031815239e-06, "loss": 0.0149, "step": 65010 }, { "grad_norm": 0.3444298803806305, "learning_rate": 9.284757856225229e-06, "loss": 0.0114, "step": 65020 }, { "grad_norm": 0.16233770549297333, "learning_rate": 9.272764637756538e-06, "loss": 0.0105, "step": 65030 }, { "grad_norm": 0.2861998677253723, "learning_rate": 9.260778378458479e-06, "loss": 0.0181, "step": 65040 }, { "grad_norm": 0.1213693767786026, "learning_rate": 9.248799080379172e-06, "loss": 0.0136, "step": 65050 }, { "grad_norm": 0.2981465756893158, "learning_rate": 9.236826745565558e-06, "loss": 0.0149, "step": 65060 }, { "grad_norm": 0.21201589703559875, "learning_rate": 9.224861376063388e-06, "loss": 0.0145, "step": 65070 }, { "grad_norm": 0.281179815530777, "learning_rate": 9.212902973917192e-06, "loss": 0.0109, "step": 65080 }, { "grad_norm": 0.2953455448150635, "learning_rate": 9.20095154117035e-06, "loss": 0.0099, "step": 65090 }, { "grad_norm": 0.30704981088638306, "learning_rate": 9.189007079865036e-06, "loss": 0.0103, "step": 65100 }, { "grad_norm": 0.542888879776001, "learning_rate": 9.177069592042226e-06, "loss": 0.0138, "step": 65110 }, { "grad_norm": 0.2310853898525238, "learning_rate": 9.165139079741724e-06, "loss": 0.0182, "step": 65120 }, { "grad_norm": 1.0829228162765503, "learning_rate": 9.153215545002098e-06, "loss": 0.0169, "step": 65130 }, { "grad_norm": 0.3867311477661133, "learning_rate": 9.141298989860798e-06, "loss": 0.0189, "step": 65140 }, { "grad_norm": 0.2995346188545227, "learning_rate": 9.129389416353994e-06, "loss": 0.0115, "step": 65150 }, { "grad_norm": 0.42024892568588257, "learning_rate": 9.11748682651673e-06, "loss": 0.0116, "step": 65160 }, { "grad_norm": 0.2614940106868744, "learning_rate": 9.105591222382837e-06, "loss": 0.0126, "step": 65170 }, { "grad_norm": 0.12344911694526672, "learning_rate": 9.093702605984915e-06, "loss": 0.0137, "step": 65180 }, { "grad_norm": 0.38260123133659363, "learning_rate": 9.081820979354455e-06, "loss": 0.0143, "step": 65190 }, { "grad_norm": 0.23144066333770752, "learning_rate": 9.069946344521663e-06, "loss": 0.0162, "step": 65200 }, { "grad_norm": 0.19323810935020447, "learning_rate": 9.058078703515598e-06, "loss": 0.0145, "step": 65210 }, { "grad_norm": 0.3646632134914398, "learning_rate": 9.046218058364125e-06, "loss": 0.0216, "step": 65220 }, { "grad_norm": 0.26837486028671265, "learning_rate": 9.034364411093893e-06, "loss": 0.0147, "step": 65230 }, { "grad_norm": 0.592243492603302, "learning_rate": 9.022517763730371e-06, "loss": 0.016, "step": 65240 }, { "grad_norm": 0.4339984059333801, "learning_rate": 9.010678118297827e-06, "loss": 0.0145, "step": 65250 }, { "grad_norm": 0.35515305399894714, "learning_rate": 8.998845476819345e-06, "loss": 0.0141, "step": 65260 }, { "grad_norm": 0.4491671919822693, "learning_rate": 8.987019841316773e-06, "loss": 0.0164, "step": 65270 }, { "grad_norm": 0.28773245215415955, "learning_rate": 8.975201213810802e-06, "loss": 0.0074, "step": 65280 }, { "grad_norm": 0.38698169589042664, "learning_rate": 8.963389596320915e-06, "loss": 0.0138, "step": 65290 }, { "grad_norm": 0.26458290219306946, "learning_rate": 8.951584990865391e-06, "loss": 0.0133, "step": 65300 }, { "grad_norm": 0.3803046643733978, "learning_rate": 8.939787399461319e-06, "loss": 0.0176, "step": 65310 }, { "grad_norm": 0.35054177045822144, "learning_rate": 8.927996824124591e-06, "loss": 0.0141, "step": 65320 }, { "grad_norm": 0.2401909977197647, "learning_rate": 8.916213266869854e-06, "loss": 0.0178, "step": 65330 }, { "grad_norm": 0.32845157384872437, "learning_rate": 8.904436729710658e-06, "loss": 0.0129, "step": 65340 }, { "grad_norm": 0.3218767046928406, "learning_rate": 8.892667214659245e-06, "loss": 0.0117, "step": 65350 }, { "grad_norm": 0.1859692484140396, "learning_rate": 8.880904723726713e-06, "loss": 0.0131, "step": 65360 }, { "grad_norm": 0.3014060854911804, "learning_rate": 8.869149258922971e-06, "loss": 0.0108, "step": 65370 }, { "grad_norm": 0.1913975477218628, "learning_rate": 8.857400822256662e-06, "loss": 0.0105, "step": 65380 }, { "grad_norm": 0.2974037528038025, "learning_rate": 8.845659415735324e-06, "loss": 0.0164, "step": 65390 }, { "grad_norm": 0.45991918444633484, "learning_rate": 8.83392504136521e-06, "loss": 0.0141, "step": 65400 }, { "grad_norm": 0.12177860736846924, "learning_rate": 8.822197701151407e-06, "loss": 0.0205, "step": 65410 }, { "grad_norm": 0.19079260528087616, "learning_rate": 8.810477397097804e-06, "loss": 0.0137, "step": 65420 }, { "grad_norm": 0.18005633354187012, "learning_rate": 8.798764131207077e-06, "loss": 0.0123, "step": 65430 }, { "grad_norm": 0.26893380284309387, "learning_rate": 8.787057905480706e-06, "loss": 0.0161, "step": 65440 }, { "grad_norm": 0.20787188410758972, "learning_rate": 8.775358721918958e-06, "loss": 0.0131, "step": 65450 }, { "grad_norm": 0.31531232595443726, "learning_rate": 8.763666582520923e-06, "loss": 0.014, "step": 65460 }, { "grad_norm": 0.20603972673416138, "learning_rate": 8.751981489284445e-06, "loss": 0.0116, "step": 65470 }, { "grad_norm": 0.08471367508172989, "learning_rate": 8.740303444206188e-06, "loss": 0.0097, "step": 65480 }, { "grad_norm": 0.18858423829078674, "learning_rate": 8.72863244928162e-06, "loss": 0.0118, "step": 65490 }, { "grad_norm": 0.26026010513305664, "learning_rate": 8.716968506504991e-06, "loss": 0.0187, "step": 65500 }, { "grad_norm": 0.39420831203460693, "learning_rate": 8.70531161786935e-06, "loss": 0.0097, "step": 65510 }, { "grad_norm": 0.19343332946300507, "learning_rate": 8.693661785366558e-06, "loss": 0.0186, "step": 65520 }, { "grad_norm": 0.2642742693424225, "learning_rate": 8.682019010987208e-06, "loss": 0.0181, "step": 65530 }, { "grad_norm": 0.44790253043174744, "learning_rate": 8.670383296720786e-06, "loss": 0.0114, "step": 65540 }, { "grad_norm": 0.2529352307319641, "learning_rate": 8.658754644555478e-06, "loss": 0.0132, "step": 65550 }, { "grad_norm": 0.2537253201007843, "learning_rate": 8.647133056478313e-06, "loss": 0.0134, "step": 65560 }, { "grad_norm": 0.22382408380508423, "learning_rate": 8.635518534475123e-06, "loss": 0.0123, "step": 65570 }, { "grad_norm": 0.4171556532382965, "learning_rate": 8.623911080530467e-06, "loss": 0.0145, "step": 65580 }, { "grad_norm": 0.11937005072832108, "learning_rate": 8.61231069662779e-06, "loss": 0.0126, "step": 65590 }, { "grad_norm": 0.33187657594680786, "learning_rate": 8.600717384749252e-06, "loss": 0.0106, "step": 65600 }, { "grad_norm": 0.29302164912223816, "learning_rate": 8.58913114687584e-06, "loss": 0.011, "step": 65610 }, { "grad_norm": 0.3410652279853821, "learning_rate": 8.57755198498732e-06, "loss": 0.0159, "step": 65620 }, { "grad_norm": 0.2733807861804962, "learning_rate": 8.565979901062265e-06, "loss": 0.018, "step": 65630 }, { "grad_norm": 0.19817210733890533, "learning_rate": 8.554414897078033e-06, "loss": 0.0146, "step": 65640 }, { "grad_norm": 0.15153768658638, "learning_rate": 8.542856975010727e-06, "loss": 0.0109, "step": 65650 }, { "grad_norm": 0.18369807302951813, "learning_rate": 8.531306136835337e-06, "loss": 0.0111, "step": 65660 }, { "grad_norm": 0.22877223789691925, "learning_rate": 8.519762384525548e-06, "loss": 0.011, "step": 65670 }, { "grad_norm": 0.24718040227890015, "learning_rate": 8.508225720053875e-06, "loss": 0.0122, "step": 65680 }, { "grad_norm": 0.2600906789302826, "learning_rate": 8.496696145391625e-06, "loss": 0.0099, "step": 65690 }, { "grad_norm": 0.25287163257598877, "learning_rate": 8.485173662508889e-06, "loss": 0.013, "step": 65700 }, { "grad_norm": 0.1994258463382721, "learning_rate": 8.473658273374536e-06, "loss": 0.0106, "step": 65710 }, { "grad_norm": 0.15723149478435516, "learning_rate": 8.462149979956253e-06, "loss": 0.0091, "step": 65720 }, { "grad_norm": 0.2331119030714035, "learning_rate": 8.450648784220461e-06, "loss": 0.0129, "step": 65730 }, { "grad_norm": 0.3292054235935211, "learning_rate": 8.439154688132417e-06, "loss": 0.0139, "step": 65740 }, { "grad_norm": 0.37808957695961, "learning_rate": 8.427667693656143e-06, "loss": 0.0167, "step": 65750 }, { "grad_norm": 0.2739730179309845, "learning_rate": 8.416187802754454e-06, "loss": 0.0129, "step": 65760 }, { "grad_norm": 0.12221517413854599, "learning_rate": 8.404715017388965e-06, "loss": 0.0102, "step": 65770 }, { "grad_norm": 0.24356518685817719, "learning_rate": 8.393249339520015e-06, "loss": 0.0123, "step": 65780 }, { "grad_norm": 0.26140835881233215, "learning_rate": 8.381790771106834e-06, "loss": 0.0104, "step": 65790 }, { "grad_norm": 0.3290582597255707, "learning_rate": 8.370339314107339e-06, "loss": 0.012, "step": 65800 }, { "grad_norm": 0.36783796548843384, "learning_rate": 8.358894970478281e-06, "loss": 0.0128, "step": 65810 }, { "grad_norm": 0.22970819473266602, "learning_rate": 8.347457742175196e-06, "loss": 0.0156, "step": 65820 }, { "grad_norm": 0.3036922216415405, "learning_rate": 8.33602763115236e-06, "loss": 0.0102, "step": 65830 }, { "grad_norm": 0.23450931906700134, "learning_rate": 8.324604639362916e-06, "loss": 0.0212, "step": 65840 }, { "grad_norm": 0.11019017547369003, "learning_rate": 8.31318876875869e-06, "loss": 0.0104, "step": 65850 }, { "grad_norm": 0.30442824959754944, "learning_rate": 8.30178002129039e-06, "loss": 0.009, "step": 65860 }, { "grad_norm": 0.20336146652698517, "learning_rate": 8.290378398907423e-06, "loss": 0.0135, "step": 65870 }, { "grad_norm": 0.24713699519634247, "learning_rate": 8.278983903558029e-06, "loss": 0.0179, "step": 65880 }, { "grad_norm": 0.4491633474826813, "learning_rate": 8.26759653718921e-06, "loss": 0.0144, "step": 65890 }, { "grad_norm": 0.18947868049144745, "learning_rate": 8.25621630174676e-06, "loss": 0.0151, "step": 65900 }, { "grad_norm": 0.46727269887924194, "learning_rate": 8.244843199175261e-06, "loss": 0.0206, "step": 65910 }, { "grad_norm": 0.20217479765415192, "learning_rate": 8.233477231418046e-06, "loss": 0.0102, "step": 65920 }, { "grad_norm": 0.19936266541481018, "learning_rate": 8.22211840041725e-06, "loss": 0.013, "step": 65930 }, { "grad_norm": 0.4099838137626648, "learning_rate": 8.210766708113792e-06, "loss": 0.0162, "step": 65940 }, { "grad_norm": 0.30012813210487366, "learning_rate": 8.199422156447367e-06, "loss": 0.0141, "step": 65950 }, { "grad_norm": 0.25662127137184143, "learning_rate": 8.188084747356451e-06, "loss": 0.0178, "step": 65960 }, { "grad_norm": 0.24022060632705688, "learning_rate": 8.176754482778299e-06, "loss": 0.0116, "step": 65970 }, { "grad_norm": 0.36873355507850647, "learning_rate": 8.165431364648918e-06, "loss": 0.0162, "step": 65980 }, { "grad_norm": 0.3413532078266144, "learning_rate": 8.154115394903162e-06, "loss": 0.0133, "step": 65990 }, { "grad_norm": 0.25453606247901917, "learning_rate": 8.142806575474582e-06, "loss": 0.0117, "step": 66000 }, { "grad_norm": 0.1694764345884323, "learning_rate": 8.131504908295562e-06, "loss": 0.0145, "step": 66010 }, { "grad_norm": 0.29424548149108887, "learning_rate": 8.120210395297262e-06, "loss": 0.0116, "step": 66020 }, { "grad_norm": 0.23025810718536377, "learning_rate": 8.108923038409565e-06, "loss": 0.0146, "step": 66030 }, { "grad_norm": 0.2903333902359009, "learning_rate": 8.097642839561226e-06, "loss": 0.0131, "step": 66040 }, { "grad_norm": 0.3265112638473511, "learning_rate": 8.08636980067966e-06, "loss": 0.0161, "step": 66050 }, { "grad_norm": 0.20115263760089874, "learning_rate": 8.075103923691186e-06, "loss": 0.0087, "step": 66060 }, { "grad_norm": 0.24628037214279175, "learning_rate": 8.063845210520793e-06, "loss": 0.0095, "step": 66070 }, { "grad_norm": 0.3433297276496887, "learning_rate": 8.052593663092295e-06, "loss": 0.0131, "step": 66080 }, { "grad_norm": 0.17550557851791382, "learning_rate": 8.04134928332827e-06, "loss": 0.0109, "step": 66090 }, { "grad_norm": 0.2077914923429489, "learning_rate": 8.030112073150086e-06, "loss": 0.0197, "step": 66100 }, { "grad_norm": 0.1529674232006073, "learning_rate": 8.018882034477881e-06, "loss": 0.0077, "step": 66110 }, { "grad_norm": 0.1942586451768875, "learning_rate": 8.007659169230541e-06, "loss": 0.0091, "step": 66120 }, { "grad_norm": 0.19884367287158966, "learning_rate": 7.996443479325755e-06, "loss": 0.0109, "step": 66130 }, { "grad_norm": 0.10359978675842285, "learning_rate": 7.985234966679977e-06, "loss": 0.0183, "step": 66140 }, { "grad_norm": 0.19527287781238556, "learning_rate": 7.974033633208438e-06, "loss": 0.0177, "step": 66150 }, { "grad_norm": 0.40877220034599304, "learning_rate": 7.962839480825135e-06, "loss": 0.0128, "step": 66160 }, { "grad_norm": 0.5023480653762817, "learning_rate": 7.951652511442858e-06, "loss": 0.0165, "step": 66170 }, { "grad_norm": 0.2813189923763275, "learning_rate": 7.940472726973125e-06, "loss": 0.0188, "step": 66180 }, { "grad_norm": 0.234388530254364, "learning_rate": 7.929300129326289e-06, "loss": 0.0136, "step": 66190 }, { "grad_norm": 0.21060609817504883, "learning_rate": 7.91813472041142e-06, "loss": 0.0126, "step": 66200 }, { "grad_norm": 0.3874264657497406, "learning_rate": 7.906976502136376e-06, "loss": 0.0122, "step": 66210 }, { "grad_norm": 0.20233014225959778, "learning_rate": 7.89582547640782e-06, "loss": 0.0113, "step": 66220 }, { "grad_norm": 0.4795299172401428, "learning_rate": 7.884681645131115e-06, "loss": 0.012, "step": 66230 }, { "grad_norm": 0.3653847277164459, "learning_rate": 7.87354501021048e-06, "loss": 0.0124, "step": 66240 }, { "grad_norm": 0.2407904863357544, "learning_rate": 7.86241557354882e-06, "loss": 0.0153, "step": 66250 }, { "grad_norm": 0.2694312036037445, "learning_rate": 7.8512933370479e-06, "loss": 0.0138, "step": 66260 }, { "grad_norm": 0.26125097274780273, "learning_rate": 7.840178302608158e-06, "loss": 0.0137, "step": 66270 }, { "grad_norm": 0.17815855145454407, "learning_rate": 7.82907047212888e-06, "loss": 0.0152, "step": 66280 }, { "grad_norm": 0.24892555177211761, "learning_rate": 7.81796984750809e-06, "loss": 0.013, "step": 66290 }, { "grad_norm": 0.19584515690803528, "learning_rate": 7.806876430642546e-06, "loss": 0.0138, "step": 66300 }, { "grad_norm": 0.207124724984169, "learning_rate": 7.795790223427862e-06, "loss": 0.0141, "step": 66310 }, { "grad_norm": 0.4077133238315582, "learning_rate": 7.784711227758324e-06, "loss": 0.0148, "step": 66320 }, { "grad_norm": 0.16869305074214935, "learning_rate": 7.773639445527053e-06, "loss": 0.0131, "step": 66330 }, { "grad_norm": 0.16308823227882385, "learning_rate": 7.762574878625905e-06, "loss": 0.0141, "step": 66340 }, { "grad_norm": 0.2742072343826294, "learning_rate": 7.751517528945513e-06, "loss": 0.0105, "step": 66350 }, { "grad_norm": 0.3345852494239807, "learning_rate": 7.740467398375278e-06, "loss": 0.013, "step": 66360 }, { "grad_norm": 0.3633795380592346, "learning_rate": 7.729424488803378e-06, "loss": 0.0105, "step": 66370 }, { "grad_norm": 0.17999301850795746, "learning_rate": 7.71838880211671e-06, "loss": 0.0199, "step": 66380 }, { "grad_norm": 0.3939211964607239, "learning_rate": 7.707360340200997e-06, "loss": 0.0119, "step": 66390 }, { "grad_norm": 0.14412830770015717, "learning_rate": 7.696339104940697e-06, "loss": 0.0109, "step": 66400 }, { "grad_norm": 0.4056795835494995, "learning_rate": 7.685325098219038e-06, "loss": 0.0152, "step": 66410 }, { "grad_norm": 0.13072189688682556, "learning_rate": 7.674318321918017e-06, "loss": 0.0133, "step": 66420 }, { "grad_norm": 0.7347712516784668, "learning_rate": 7.663318777918366e-06, "loss": 0.0189, "step": 66430 }, { "grad_norm": 0.32574355602264404, "learning_rate": 7.652326468099647e-06, "loss": 0.0159, "step": 66440 }, { "grad_norm": 0.3694325387477875, "learning_rate": 7.641341394340096e-06, "loss": 0.0125, "step": 66450 }, { "grad_norm": 0.4040897488594055, "learning_rate": 7.630363558516818e-06, "loss": 0.012, "step": 66460 }, { "grad_norm": 0.4294373095035553, "learning_rate": 7.619392962505578e-06, "loss": 0.0166, "step": 66470 }, { "grad_norm": 0.27142786979675293, "learning_rate": 7.6084296081809725e-06, "loss": 0.0188, "step": 66480 }, { "grad_norm": 0.2799489200115204, "learning_rate": 7.597473497416347e-06, "loss": 0.0155, "step": 66490 }, { "grad_norm": 0.2531890869140625, "learning_rate": 7.586524632083764e-06, "loss": 0.0131, "step": 66500 }, { "grad_norm": 0.19564568996429443, "learning_rate": 7.5755830140541326e-06, "loss": 0.014, "step": 66510 }, { "grad_norm": 0.32526618242263794, "learning_rate": 7.5646486451970475e-06, "loss": 0.0127, "step": 66520 }, { "grad_norm": 0.7208928465843201, "learning_rate": 7.553721527380897e-06, "loss": 0.0114, "step": 66530 }, { "grad_norm": 0.20555858314037323, "learning_rate": 7.542801662472826e-06, "loss": 0.012, "step": 66540 }, { "grad_norm": 0.24916759133338928, "learning_rate": 7.5318890523387475e-06, "loss": 0.0153, "step": 66550 }, { "grad_norm": 0.23120811581611633, "learning_rate": 7.52098369884332e-06, "loss": 0.0146, "step": 66560 }, { "grad_norm": 0.17884598672389984, "learning_rate": 7.510085603849992e-06, "loss": 0.0115, "step": 66570 }, { "grad_norm": 0.357746422290802, "learning_rate": 7.499194769220918e-06, "loss": 0.0131, "step": 66580 }, { "grad_norm": 0.5442632436752319, "learning_rate": 7.48831119681706e-06, "loss": 0.0178, "step": 66590 }, { "grad_norm": 0.3334491550922394, "learning_rate": 7.477434888498119e-06, "loss": 0.0089, "step": 66600 }, { "grad_norm": 0.2063235640525818, "learning_rate": 7.466565846122564e-06, "loss": 0.0223, "step": 66610 }, { "grad_norm": 0.3298609256744385, "learning_rate": 7.455704071547626e-06, "loss": 0.0132, "step": 66620 }, { "grad_norm": 0.16680458188056946, "learning_rate": 7.444849566629247e-06, "loss": 0.0126, "step": 66630 }, { "grad_norm": 0.47487202286720276, "learning_rate": 7.434002333222212e-06, "loss": 0.0116, "step": 66640 }, { "grad_norm": 0.31501835584640503, "learning_rate": 7.423162373179976e-06, "loss": 0.0153, "step": 66650 }, { "grad_norm": 0.3278263807296753, "learning_rate": 7.412329688354835e-06, "loss": 0.0134, "step": 66660 }, { "grad_norm": 0.17819298803806305, "learning_rate": 7.40150428059776e-06, "loss": 0.0112, "step": 66670 }, { "grad_norm": 0.39396238327026367, "learning_rate": 7.3906861517585354e-06, "loss": 0.0164, "step": 66680 }, { "grad_norm": 0.3029542565345764, "learning_rate": 7.37987530368569e-06, "loss": 0.0152, "step": 66690 }, { "grad_norm": 0.23705537617206573, "learning_rate": 7.369071738226474e-06, "loss": 0.0131, "step": 66700 }, { "grad_norm": 0.2457246482372284, "learning_rate": 7.358275457226954e-06, "loss": 0.0148, "step": 66710 }, { "grad_norm": 0.2494760900735855, "learning_rate": 7.347486462531899e-06, "loss": 0.0108, "step": 66720 }, { "grad_norm": 0.1608937382698059, "learning_rate": 7.336704755984858e-06, "loss": 0.0082, "step": 66730 }, { "grad_norm": 0.19265136122703552, "learning_rate": 7.325930339428133e-06, "loss": 0.0153, "step": 66740 }, { "grad_norm": 0.28080686926841736, "learning_rate": 7.315163214702769e-06, "loss": 0.0102, "step": 66750 }, { "grad_norm": 0.23131002485752106, "learning_rate": 7.304403383648595e-06, "loss": 0.0104, "step": 66760 }, { "grad_norm": 0.133949413895607, "learning_rate": 7.293650848104139e-06, "loss": 0.0147, "step": 66770 }, { "grad_norm": 0.18669484555721283, "learning_rate": 7.2829056099067374e-06, "loss": 0.0141, "step": 66780 }, { "grad_norm": 0.2480178028345108, "learning_rate": 7.2721676708924494e-06, "loss": 0.0133, "step": 66790 }, { "grad_norm": 0.4938141703605652, "learning_rate": 7.261437032896096e-06, "loss": 0.0216, "step": 66800 }, { "grad_norm": 0.1970716267824173, "learning_rate": 7.250713697751255e-06, "loss": 0.0133, "step": 66810 }, { "grad_norm": 0.18945132195949554, "learning_rate": 7.239997667290255e-06, "loss": 0.0094, "step": 66820 }, { "grad_norm": 0.15851466357707977, "learning_rate": 7.2292889433441425e-06, "loss": 0.0117, "step": 66830 }, { "grad_norm": 0.29417890310287476, "learning_rate": 7.218587527742793e-06, "loss": 0.0112, "step": 66840 }, { "grad_norm": 0.2772735059261322, "learning_rate": 7.207893422314749e-06, "loss": 0.011, "step": 66850 }, { "grad_norm": 0.3309766352176666, "learning_rate": 7.1972066288873545e-06, "loss": 0.0103, "step": 66860 }, { "grad_norm": 0.30711841583251953, "learning_rate": 7.186527149286687e-06, "loss": 0.0222, "step": 66870 }, { "grad_norm": 0.20955505967140198, "learning_rate": 7.175854985337576e-06, "loss": 0.0128, "step": 66880 }, { "grad_norm": 0.23162417113780975, "learning_rate": 7.16519013886362e-06, "loss": 0.015, "step": 66890 }, { "grad_norm": 0.20099574327468872, "learning_rate": 7.154532611687109e-06, "loss": 0.0099, "step": 66900 }, { "grad_norm": 0.18037454783916473, "learning_rate": 7.143882405629177e-06, "loss": 0.0121, "step": 66910 }, { "grad_norm": 0.3042151629924774, "learning_rate": 7.133239522509605e-06, "loss": 0.0112, "step": 66920 }, { "grad_norm": 0.14354561269283295, "learning_rate": 7.1226039641469956e-06, "loss": 0.0117, "step": 66930 }, { "grad_norm": 0.24332402646541595, "learning_rate": 7.111975732358678e-06, "loss": 0.0112, "step": 66940 }, { "grad_norm": 0.17534443736076355, "learning_rate": 7.101354828960693e-06, "loss": 0.011, "step": 66950 }, { "grad_norm": 0.44159916043281555, "learning_rate": 7.090741255767918e-06, "loss": 0.0125, "step": 66960 }, { "grad_norm": 0.2276603877544403, "learning_rate": 7.080135014593875e-06, "loss": 0.013, "step": 66970 }, { "grad_norm": 0.20729117095470428, "learning_rate": 7.069536107250896e-06, "loss": 0.0154, "step": 66980 }, { "grad_norm": 0.07322744280099869, "learning_rate": 7.058944535550049e-06, "loss": 0.0102, "step": 66990 }, { "grad_norm": 0.3050169348716736, "learning_rate": 7.048360301301138e-06, "loss": 0.0127, "step": 67000 }, { "grad_norm": 0.33213210105895996, "learning_rate": 7.03778340631272e-06, "loss": 0.0137, "step": 67010 }, { "grad_norm": 0.28174784779548645, "learning_rate": 7.0272138523921e-06, "loss": 0.0136, "step": 67020 }, { "grad_norm": 0.15645205974578857, "learning_rate": 7.016651641345334e-06, "loss": 0.0135, "step": 67030 }, { "grad_norm": 0.2753855288028717, "learning_rate": 7.0060967749772e-06, "loss": 0.0105, "step": 67040 }, { "grad_norm": 0.4335983991622925, "learning_rate": 6.995549255091238e-06, "loss": 0.0218, "step": 67050 }, { "grad_norm": 0.22334879636764526, "learning_rate": 6.98500908348973e-06, "loss": 0.0171, "step": 67060 }, { "grad_norm": 0.6295152902603149, "learning_rate": 6.974476261973711e-06, "loss": 0.0179, "step": 67070 }, { "grad_norm": 0.14797554910182953, "learning_rate": 6.963950792342949e-06, "loss": 0.0091, "step": 67080 }, { "grad_norm": 0.19626182317733765, "learning_rate": 6.9534326763959715e-06, "loss": 0.0105, "step": 67090 }, { "grad_norm": 0.31666576862335205, "learning_rate": 6.942921915929995e-06, "loss": 0.0163, "step": 67100 }, { "grad_norm": 0.30013957619667053, "learning_rate": 6.9324185127410735e-06, "loss": 0.0105, "step": 67110 }, { "grad_norm": 0.2117512971162796, "learning_rate": 6.921922468623921e-06, "loss": 0.0113, "step": 67120 }, { "grad_norm": 0.13813073933124542, "learning_rate": 6.911433785372023e-06, "loss": 0.0129, "step": 67130 }, { "grad_norm": 0.31811654567718506, "learning_rate": 6.900952464777632e-06, "loss": 0.014, "step": 67140 }, { "grad_norm": 0.24371686577796936, "learning_rate": 6.8904785086316815e-06, "loss": 0.015, "step": 67150 }, { "grad_norm": 0.27780163288116455, "learning_rate": 6.880011918723927e-06, "loss": 0.0142, "step": 67160 }, { "grad_norm": 0.34251853823661804, "learning_rate": 6.86955269684279e-06, "loss": 0.0106, "step": 67170 }, { "grad_norm": 0.29896625876426697, "learning_rate": 6.859100844775473e-06, "loss": 0.0203, "step": 67180 }, { "grad_norm": 0.2584563195705414, "learning_rate": 6.84865636430792e-06, "loss": 0.0135, "step": 67190 }, { "grad_norm": 0.1965472251176834, "learning_rate": 6.838219257224804e-06, "loss": 0.0097, "step": 67200 }, { "grad_norm": 0.12236601114273071, "learning_rate": 6.827789525309536e-06, "loss": 0.015, "step": 67210 }, { "grad_norm": 0.3229636549949646, "learning_rate": 6.81736717034428e-06, "loss": 0.0154, "step": 67220 }, { "grad_norm": 0.12905466556549072, "learning_rate": 6.806952194109933e-06, "loss": 0.0193, "step": 67230 }, { "grad_norm": 0.5218296051025391, "learning_rate": 6.796544598386112e-06, "loss": 0.0084, "step": 67240 }, { "grad_norm": 0.2873811423778534, "learning_rate": 6.786144384951204e-06, "loss": 0.0137, "step": 67250 }, { "grad_norm": 0.28544580936431885, "learning_rate": 6.775751555582322e-06, "loss": 0.0117, "step": 67260 }, { "grad_norm": 0.28883087635040283, "learning_rate": 6.76536611205531e-06, "loss": 0.0131, "step": 67270 }, { "grad_norm": 0.15294678509235382, "learning_rate": 6.754988056144762e-06, "loss": 0.0143, "step": 67280 }, { "grad_norm": 0.356845498085022, "learning_rate": 6.744617389624014e-06, "loss": 0.0143, "step": 67290 }, { "grad_norm": 0.3735763728618622, "learning_rate": 6.734254114265087e-06, "loss": 0.0132, "step": 67300 }, { "grad_norm": 0.4826510548591614, "learning_rate": 6.723898231838843e-06, "loss": 0.0106, "step": 67310 }, { "grad_norm": 0.33850333094596863, "learning_rate": 6.713549744114766e-06, "loss": 0.0157, "step": 67320 }, { "grad_norm": 0.2591692805290222, "learning_rate": 6.703208652861159e-06, "loss": 0.012, "step": 67330 }, { "grad_norm": 0.4026529788970947, "learning_rate": 6.69287495984503e-06, "loss": 0.0138, "step": 67340 }, { "grad_norm": 0.33459848165512085, "learning_rate": 6.6825486668321e-06, "loss": 0.0155, "step": 67350 }, { "grad_norm": 0.22572526335716248, "learning_rate": 6.672229775586886e-06, "loss": 0.0128, "step": 67360 }, { "grad_norm": 0.36599814891815186, "learning_rate": 6.661918287872576e-06, "loss": 0.0116, "step": 67370 }, { "grad_norm": 0.48827633261680603, "learning_rate": 6.6516142054511346e-06, "loss": 0.0142, "step": 67380 }, { "grad_norm": 0.21380820870399475, "learning_rate": 6.641317530083241e-06, "loss": 0.0114, "step": 67390 }, { "grad_norm": 0.14737555384635925, "learning_rate": 6.631028263528322e-06, "loss": 0.015, "step": 67400 }, { "grad_norm": 0.2481212019920349, "learning_rate": 6.620746407544537e-06, "loss": 0.0119, "step": 67410 }, { "grad_norm": 0.18608181178569794, "learning_rate": 6.610471963888742e-06, "loss": 0.0113, "step": 67420 }, { "grad_norm": 0.18919886648654938, "learning_rate": 6.600204934316606e-06, "loss": 0.0111, "step": 67430 }, { "grad_norm": 0.3187160789966583, "learning_rate": 6.589945320582452e-06, "loss": 0.0125, "step": 67440 }, { "grad_norm": 0.2614441514015198, "learning_rate": 6.579693124439374e-06, "loss": 0.0128, "step": 67450 }, { "grad_norm": 0.27605926990509033, "learning_rate": 6.56944834763919e-06, "loss": 0.0098, "step": 67460 }, { "grad_norm": 0.275320827960968, "learning_rate": 6.5592109919324575e-06, "loss": 0.0114, "step": 67470 }, { "grad_norm": 0.5070396661758423, "learning_rate": 6.54898105906846e-06, "loss": 0.0128, "step": 67480 }, { "grad_norm": 0.13229985535144806, "learning_rate": 6.5387585507952155e-06, "loss": 0.0202, "step": 67490 }, { "grad_norm": 0.3204885423183441, "learning_rate": 6.528543468859461e-06, "loss": 0.0113, "step": 67500 }, { "grad_norm": 0.35451391339302063, "learning_rate": 6.518335815006682e-06, "loss": 0.0171, "step": 67510 }, { "grad_norm": 0.26202648878097534, "learning_rate": 6.5081355909810845e-06, "loss": 0.0134, "step": 67520 }, { "grad_norm": 0.22191192209720612, "learning_rate": 6.497942798525608e-06, "loss": 0.018, "step": 67530 }, { "grad_norm": 0.46101129055023193, "learning_rate": 6.487757439381936e-06, "loss": 0.0183, "step": 67540 }, { "grad_norm": 0.17991256713867188, "learning_rate": 6.477579515290433e-06, "loss": 0.0161, "step": 67550 }, { "grad_norm": 0.20398356020450592, "learning_rate": 6.467409027990273e-06, "loss": 0.0154, "step": 67560 }, { "grad_norm": 0.13045011460781097, "learning_rate": 6.457245979219279e-06, "loss": 0.0202, "step": 67570 }, { "grad_norm": 0.3161429464817047, "learning_rate": 6.447090370714054e-06, "loss": 0.02, "step": 67580 }, { "grad_norm": 0.11774042993783951, "learning_rate": 6.436942204209917e-06, "loss": 0.0167, "step": 67590 }, { "grad_norm": 0.2791283428668976, "learning_rate": 6.4268014814408804e-06, "loss": 0.0159, "step": 67600 }, { "grad_norm": 0.1998234987258911, "learning_rate": 6.41666820413977e-06, "loss": 0.0122, "step": 67610 }, { "grad_norm": 0.3273558020591736, "learning_rate": 6.406542374038033e-06, "loss": 0.017, "step": 67620 }, { "grad_norm": 0.23687663674354553, "learning_rate": 6.396423992865935e-06, "loss": 0.0096, "step": 67630 }, { "grad_norm": 0.24716301262378693, "learning_rate": 6.386313062352412e-06, "loss": 0.0129, "step": 67640 }, { "grad_norm": 0.2438616007566452, "learning_rate": 6.376209584225152e-06, "loss": 0.0091, "step": 67650 }, { "grad_norm": 0.5410526990890503, "learning_rate": 6.366113560210557e-06, "loss": 0.0106, "step": 67660 }, { "grad_norm": 0.4622226357460022, "learning_rate": 6.356024992033766e-06, "loss": 0.0134, "step": 67670 }, { "grad_norm": 0.374599814414978, "learning_rate": 6.345943881418648e-06, "loss": 0.0117, "step": 67680 }, { "grad_norm": 0.22330531477928162, "learning_rate": 6.33587023008777e-06, "loss": 0.0107, "step": 67690 }, { "grad_norm": 0.15451723337173462, "learning_rate": 6.32580403976245e-06, "loss": 0.0103, "step": 67700 }, { "grad_norm": 0.3463292717933655, "learning_rate": 6.31574531216273e-06, "loss": 0.0145, "step": 67710 }, { "grad_norm": 0.26614055037498474, "learning_rate": 6.305694049007371e-06, "loss": 0.0161, "step": 67720 }, { "grad_norm": 0.3221202492713928, "learning_rate": 6.2956502520138575e-06, "loss": 0.0129, "step": 67730 }, { "grad_norm": 0.24232596158981323, "learning_rate": 6.285613922898409e-06, "loss": 0.01, "step": 67740 }, { "grad_norm": 0.1838638037443161, "learning_rate": 6.275585063375927e-06, "loss": 0.0165, "step": 67750 }, { "grad_norm": 0.0877019390463829, "learning_rate": 6.265563675160113e-06, "loss": 0.0137, "step": 67760 }, { "grad_norm": 0.14487572014331818, "learning_rate": 6.25554975996332e-06, "loss": 0.0117, "step": 67770 }, { "grad_norm": 0.28517574071884155, "learning_rate": 6.245543319496661e-06, "loss": 0.0099, "step": 67780 }, { "grad_norm": 0.23283572494983673, "learning_rate": 6.2355443554699685e-06, "loss": 0.0155, "step": 67790 }, { "grad_norm": 0.41768836975097656, "learning_rate": 6.225552869591766e-06, "loss": 0.0108, "step": 67800 }, { "grad_norm": 0.2065265029668808, "learning_rate": 6.215568863569365e-06, "loss": 0.0125, "step": 67810 }, { "grad_norm": 0.3797850012779236, "learning_rate": 6.205592339108712e-06, "loss": 0.018, "step": 67820 }, { "grad_norm": 0.32394811511039734, "learning_rate": 6.195623297914577e-06, "loss": 0.0103, "step": 67830 }, { "grad_norm": 0.2469169646501541, "learning_rate": 6.185661741690357e-06, "loss": 0.0133, "step": 67840 }, { "grad_norm": 0.30097049474716187, "learning_rate": 6.1757076721382145e-06, "loss": 0.0167, "step": 67850 }, { "grad_norm": 0.1989172250032425, "learning_rate": 6.165761090959038e-06, "loss": 0.0146, "step": 67860 }, { "grad_norm": 0.3255784213542938, "learning_rate": 6.155821999852424e-06, "loss": 0.0121, "step": 67870 }, { "grad_norm": 0.16575399041175842, "learning_rate": 6.145890400516696e-06, "loss": 0.0124, "step": 67880 }, { "grad_norm": 0.26019003987312317, "learning_rate": 6.1359662946488816e-06, "loss": 0.0135, "step": 67890 }, { "grad_norm": 0.29921144247055054, "learning_rate": 6.12604968394474e-06, "loss": 0.0096, "step": 67900 }, { "grad_norm": 0.3283664286136627, "learning_rate": 6.11614057009875e-06, "loss": 0.0144, "step": 67910 }, { "grad_norm": 0.24707743525505066, "learning_rate": 6.106238954804111e-06, "loss": 0.0108, "step": 67920 }, { "grad_norm": 0.24737553298473358, "learning_rate": 6.096344839752738e-06, "loss": 0.0165, "step": 67930 }, { "grad_norm": 0.18503667414188385, "learning_rate": 6.086458226635278e-06, "loss": 0.0119, "step": 67940 }, { "grad_norm": 0.2029855102300644, "learning_rate": 6.076579117141046e-06, "loss": 0.0153, "step": 67950 }, { "grad_norm": 0.23405346274375916, "learning_rate": 6.066707512958153e-06, "loss": 0.0096, "step": 67960 }, { "grad_norm": 0.27084991335868835, "learning_rate": 6.056843415773361e-06, "loss": 0.0173, "step": 67970 }, { "grad_norm": 0.2699402868747711, "learning_rate": 6.0469868272721776e-06, "loss": 0.0162, "step": 67980 }, { "grad_norm": 0.4301266372203827, "learning_rate": 6.037137749138844e-06, "loss": 0.0157, "step": 67990 }, { "grad_norm": 0.3669259250164032, "learning_rate": 6.027296183056252e-06, "loss": 0.0163, "step": 68000 }, { "grad_norm": 0.2894023656845093, "learning_rate": 6.017462130706114e-06, "loss": 0.0114, "step": 68010 }, { "grad_norm": 0.19390489161014557, "learning_rate": 6.0076355937687516e-06, "loss": 0.0161, "step": 68020 }, { "grad_norm": 0.16944094002246857, "learning_rate": 5.9978165739232925e-06, "loss": 0.0138, "step": 68030 }, { "grad_norm": 0.24176868796348572, "learning_rate": 5.988005072847508e-06, "loss": 0.012, "step": 68040 }, { "grad_norm": 0.2571680247783661, "learning_rate": 5.978201092217928e-06, "loss": 0.0221, "step": 68050 }, { "grad_norm": 0.16713805496692657, "learning_rate": 5.9684046337097895e-06, "loss": 0.0089, "step": 68060 }, { "grad_norm": 0.10892514884471893, "learning_rate": 5.958615698997016e-06, "loss": 0.0141, "step": 68070 }, { "grad_norm": 0.15115518867969513, "learning_rate": 5.948834289752303e-06, "loss": 0.0115, "step": 68080 }, { "grad_norm": 0.22803421318531036, "learning_rate": 5.939060407646996e-06, "loss": 0.0174, "step": 68090 }, { "grad_norm": 0.3223370909690857, "learning_rate": 5.929294054351198e-06, "loss": 0.0124, "step": 68100 }, { "grad_norm": 0.3700934648513794, "learning_rate": 5.919535231533707e-06, "loss": 0.0128, "step": 68110 }, { "grad_norm": 0.3036673367023468, "learning_rate": 5.909783940862046e-06, "loss": 0.0159, "step": 68120 }, { "grad_norm": 0.24186401069164276, "learning_rate": 5.900040184002436e-06, "loss": 0.0101, "step": 68130 }, { "grad_norm": 0.46308061480522156, "learning_rate": 5.890303962619831e-06, "loss": 0.0109, "step": 68140 }, { "grad_norm": 0.30077052116394043, "learning_rate": 5.88057527837787e-06, "loss": 0.0145, "step": 68150 }, { "grad_norm": 0.17647647857666016, "learning_rate": 5.870854132938924e-06, "loss": 0.0157, "step": 68160 }, { "grad_norm": 0.2499355524778366, "learning_rate": 5.861140527964071e-06, "loss": 0.0193, "step": 68170 }, { "grad_norm": 0.12182323634624481, "learning_rate": 5.851434465113098e-06, "loss": 0.0097, "step": 68180 }, { "grad_norm": 0.22361117601394653, "learning_rate": 5.841735946044524e-06, "loss": 0.0142, "step": 68190 }, { "grad_norm": 0.11519971489906311, "learning_rate": 5.832044972415523e-06, "loss": 0.0116, "step": 68200 }, { "grad_norm": 0.18836572766304016, "learning_rate": 5.8223615458820655e-06, "loss": 0.0152, "step": 68210 }, { "grad_norm": 0.10345585644245148, "learning_rate": 5.812685668098733e-06, "loss": 0.0126, "step": 68220 }, { "grad_norm": 0.1798819899559021, "learning_rate": 5.803017340718913e-06, "loss": 0.0141, "step": 68230 }, { "grad_norm": 0.39400047063827515, "learning_rate": 5.793356565394636e-06, "loss": 0.0124, "step": 68240 }, { "grad_norm": 0.2699209153652191, "learning_rate": 5.783703343776658e-06, "loss": 0.0105, "step": 68250 }, { "grad_norm": 0.16631899774074554, "learning_rate": 5.774057677514477e-06, "loss": 0.0111, "step": 68260 }, { "grad_norm": 0.1263572722673416, "learning_rate": 5.764419568256235e-06, "loss": 0.0145, "step": 68270 }, { "grad_norm": 0.40899157524108887, "learning_rate": 5.754789017648865e-06, "loss": 0.019, "step": 68280 }, { "grad_norm": 0.18370257318019867, "learning_rate": 5.745166027337934e-06, "loss": 0.0193, "step": 68290 }, { "grad_norm": 0.13952015340328217, "learning_rate": 5.735550598967754e-06, "loss": 0.0154, "step": 68300 }, { "grad_norm": 0.20784799754619598, "learning_rate": 5.725942734181339e-06, "loss": 0.0128, "step": 68310 }, { "grad_norm": 0.32729923725128174, "learning_rate": 5.7163424346204146e-06, "loss": 0.0126, "step": 68320 }, { "grad_norm": 0.2655428349971771, "learning_rate": 5.706749701925407e-06, "loss": 0.0115, "step": 68330 }, { "grad_norm": 0.19527702033519745, "learning_rate": 5.697164537735461e-06, "loss": 0.0101, "step": 68340 }, { "grad_norm": 0.22406119108200073, "learning_rate": 5.687586943688406e-06, "loss": 0.0076, "step": 68350 }, { "grad_norm": 0.16588136553764343, "learning_rate": 5.678016921420792e-06, "loss": 0.0118, "step": 68360 }, { "grad_norm": 0.30133306980133057, "learning_rate": 5.668454472567875e-06, "loss": 0.0223, "step": 68370 }, { "grad_norm": 0.34549447894096375, "learning_rate": 5.658899598763617e-06, "loss": 0.0141, "step": 68380 }, { "grad_norm": 0.23674440383911133, "learning_rate": 5.6493523016407035e-06, "loss": 0.0093, "step": 68390 }, { "grad_norm": 0.281797856092453, "learning_rate": 5.6398125828304615e-06, "loss": 0.0122, "step": 68400 }, { "grad_norm": 0.33558669686317444, "learning_rate": 5.630280443963015e-06, "loss": 0.0127, "step": 68410 }, { "grad_norm": 0.2621694803237915, "learning_rate": 5.620755886667112e-06, "loss": 0.0132, "step": 68420 }, { "grad_norm": 0.14524555206298828, "learning_rate": 5.611238912570271e-06, "loss": 0.0126, "step": 68430 }, { "grad_norm": 0.27961716055870056, "learning_rate": 5.601729523298649e-06, "loss": 0.0168, "step": 68440 }, { "grad_norm": 0.24953073263168335, "learning_rate": 5.592227720477161e-06, "loss": 0.0115, "step": 68450 }, { "grad_norm": 0.17651815712451935, "learning_rate": 5.582733505729415e-06, "loss": 0.0154, "step": 68460 }, { "grad_norm": 0.2542335093021393, "learning_rate": 5.573246880677668e-06, "loss": 0.0121, "step": 68470 }, { "grad_norm": 1.084369421005249, "learning_rate": 5.56376784694298e-06, "loss": 0.0129, "step": 68480 }, { "grad_norm": 0.38645297288894653, "learning_rate": 5.554296406145027e-06, "loss": 0.0145, "step": 68490 }, { "grad_norm": 0.2554646134376526, "learning_rate": 5.544832559902219e-06, "loss": 0.0127, "step": 68500 }, { "grad_norm": 0.2725629210472107, "learning_rate": 5.53537630983168e-06, "loss": 0.0138, "step": 68510 }, { "grad_norm": 0.1879866123199463, "learning_rate": 5.5259276575492125e-06, "loss": 0.016, "step": 68520 }, { "grad_norm": 0.26444312930107117, "learning_rate": 5.516486604669357e-06, "loss": 0.0167, "step": 68530 }, { "grad_norm": 0.215643510222435, "learning_rate": 5.507053152805303e-06, "loss": 0.0149, "step": 68540 }, { "grad_norm": 0.2624824047088623, "learning_rate": 5.497627303568975e-06, "loss": 0.0166, "step": 68550 }, { "grad_norm": 0.22348864376544952, "learning_rate": 5.488209058571003e-06, "loss": 0.0079, "step": 68560 }, { "grad_norm": 0.2821292281150818, "learning_rate": 5.4787984194207054e-06, "loss": 0.016, "step": 68570 }, { "grad_norm": 0.23657625913619995, "learning_rate": 5.469395387726095e-06, "loss": 0.0111, "step": 68580 }, { "grad_norm": 0.26236197352409363, "learning_rate": 5.45999996509391e-06, "loss": 0.0122, "step": 68590 }, { "grad_norm": 0.2611541748046875, "learning_rate": 5.450612153129536e-06, "loss": 0.013, "step": 68600 }, { "grad_norm": 0.21751751005649567, "learning_rate": 5.4412319534371426e-06, "loss": 0.0145, "step": 68610 }, { "grad_norm": 0.2488076239824295, "learning_rate": 5.431859367619513e-06, "loss": 0.0098, "step": 68620 }, { "grad_norm": 0.22325395047664642, "learning_rate": 5.422494397278172e-06, "loss": 0.013, "step": 68630 }, { "grad_norm": 0.4067608714103699, "learning_rate": 5.413137044013344e-06, "loss": 0.0158, "step": 68640 }, { "grad_norm": 0.31780433654785156, "learning_rate": 5.403787309423941e-06, "loss": 0.0175, "step": 68650 }, { "grad_norm": 0.14293836057186127, "learning_rate": 5.3944451951075835e-06, "loss": 0.0068, "step": 68660 }, { "grad_norm": 0.2676987648010254, "learning_rate": 5.385110702660562e-06, "loss": 0.0106, "step": 68670 }, { "grad_norm": 0.19617655873298645, "learning_rate": 5.375783833677922e-06, "loss": 0.0122, "step": 68680 }, { "grad_norm": 0.2523968815803528, "learning_rate": 5.36646458975334e-06, "loss": 0.0118, "step": 68690 }, { "grad_norm": 0.5526746511459351, "learning_rate": 5.3571529724792294e-06, "loss": 0.0113, "step": 68700 }, { "grad_norm": 0.40095919370651245, "learning_rate": 5.347848983446702e-06, "loss": 0.0145, "step": 68710 }, { "grad_norm": 0.15200500190258026, "learning_rate": 5.3385526242455185e-06, "loss": 0.0122, "step": 68720 }, { "grad_norm": 0.3661642074584961, "learning_rate": 5.329263896464226e-06, "loss": 0.0093, "step": 68730 }, { "grad_norm": 0.1304672509431839, "learning_rate": 5.3199828016899715e-06, "loss": 0.0099, "step": 68740 }, { "grad_norm": 0.29557672142982483, "learning_rate": 5.310709341508657e-06, "loss": 0.0118, "step": 68750 }, { "grad_norm": 0.2163064032793045, "learning_rate": 5.301443517504861e-06, "loss": 0.0105, "step": 68760 }, { "grad_norm": 0.1545100063085556, "learning_rate": 5.292185331261862e-06, "loss": 0.0085, "step": 68770 }, { "grad_norm": 0.2004951387643814, "learning_rate": 5.28293478436162e-06, "loss": 0.0122, "step": 68780 }, { "grad_norm": 0.7032880187034607, "learning_rate": 5.273691878384829e-06, "loss": 0.0147, "step": 68790 }, { "grad_norm": 0.2568909525871277, "learning_rate": 5.264456614910812e-06, "loss": 0.0133, "step": 68800 }, { "grad_norm": 0.32243528962135315, "learning_rate": 5.255228995517647e-06, "loss": 0.0214, "step": 68810 }, { "grad_norm": 0.163865864276886, "learning_rate": 5.246009021782067e-06, "loss": 0.0108, "step": 68820 }, { "grad_norm": 0.25739434361457825, "learning_rate": 5.2367966952795225e-06, "loss": 0.0189, "step": 68830 }, { "grad_norm": 0.21546170115470886, "learning_rate": 5.2275920175841485e-06, "loss": 0.0125, "step": 68840 }, { "grad_norm": 0.24938558042049408, "learning_rate": 5.218394990268771e-06, "loss": 0.0121, "step": 68850 }, { "grad_norm": 0.3569282591342926, "learning_rate": 5.209205614904916e-06, "loss": 0.0149, "step": 68860 }, { "grad_norm": 0.2392604649066925, "learning_rate": 5.200023893062777e-06, "loss": 0.0096, "step": 68870 }, { "grad_norm": 0.1689765602350235, "learning_rate": 5.190849826311289e-06, "loss": 0.0107, "step": 68880 }, { "grad_norm": 0.35333168506622314, "learning_rate": 5.181683416218025e-06, "loss": 0.0166, "step": 68890 }, { "grad_norm": 0.17938366532325745, "learning_rate": 5.172524664349276e-06, "loss": 0.0107, "step": 68900 }, { "grad_norm": 0.34111741185188293, "learning_rate": 5.1633735722700416e-06, "loss": 0.0163, "step": 68910 }, { "grad_norm": 0.21198558807373047, "learning_rate": 5.154230141543958e-06, "loss": 0.0116, "step": 68920 }, { "grad_norm": 0.27596771717071533, "learning_rate": 5.145094373733433e-06, "loss": 0.0126, "step": 68930 }, { "grad_norm": 0.14169202744960785, "learning_rate": 5.135966270399478e-06, "loss": 0.0103, "step": 68940 }, { "grad_norm": 0.6274378299713135, "learning_rate": 5.126845833101857e-06, "loss": 0.0151, "step": 68950 }, { "grad_norm": 0.18820123374462128, "learning_rate": 5.1177330633989994e-06, "loss": 0.0137, "step": 68960 }, { "grad_norm": 0.32736343145370483, "learning_rate": 5.108627962848033e-06, "loss": 0.0283, "step": 68970 }, { "grad_norm": 0.24801315367221832, "learning_rate": 5.099530533004759e-06, "loss": 0.0169, "step": 68980 }, { "grad_norm": 0.1985962688922882, "learning_rate": 5.090440775423699e-06, "loss": 0.0122, "step": 68990 }, { "grad_norm": 0.18815167248249054, "learning_rate": 5.081358691658022e-06, "loss": 0.013, "step": 69000 }, { "grad_norm": 0.866508960723877, "learning_rate": 5.072284283259621e-06, "loss": 0.0162, "step": 69010 }, { "grad_norm": 0.23374450206756592, "learning_rate": 5.063217551779053e-06, "loss": 0.0177, "step": 69020 }, { "grad_norm": 0.21886463463306427, "learning_rate": 5.054158498765588e-06, "loss": 0.0104, "step": 69030 }, { "grad_norm": 0.35326331853866577, "learning_rate": 5.0451071257671625e-06, "loss": 0.0132, "step": 69040 }, { "grad_norm": 0.5151524543762207, "learning_rate": 5.036063434330407e-06, "loss": 0.0196, "step": 69050 }, { "grad_norm": 0.17906856536865234, "learning_rate": 5.027027426000652e-06, "loss": 0.0134, "step": 69060 }, { "grad_norm": 0.21890556812286377, "learning_rate": 5.017999102321886e-06, "loss": 0.0139, "step": 69070 }, { "grad_norm": 0.3291625678539276, "learning_rate": 5.0089784648368224e-06, "loss": 0.0136, "step": 69080 }, { "grad_norm": 0.1423223912715912, "learning_rate": 4.999965515086829e-06, "loss": 0.0092, "step": 69090 }, { "grad_norm": 0.2004816234111786, "learning_rate": 4.990960254611976e-06, "loss": 0.0109, "step": 69100 }, { "grad_norm": 0.36048269271850586, "learning_rate": 4.9819626849510194e-06, "loss": 0.0155, "step": 69110 }, { "grad_norm": 0.21465784311294556, "learning_rate": 4.97297280764138e-06, "loss": 0.0166, "step": 69120 }, { "grad_norm": 0.18070866167545319, "learning_rate": 4.963990624219211e-06, "loss": 0.0138, "step": 69130 }, { "grad_norm": 0.3234412372112274, "learning_rate": 4.955016136219298e-06, "loss": 0.0131, "step": 69140 }, { "grad_norm": 0.19740308821201324, "learning_rate": 4.946049345175146e-06, "loss": 0.0103, "step": 69150 }, { "grad_norm": 0.35855334997177124, "learning_rate": 4.937090252618937e-06, "loss": 0.0143, "step": 69160 }, { "grad_norm": 0.3585274815559387, "learning_rate": 4.928138860081521e-06, "loss": 0.0161, "step": 69170 }, { "grad_norm": 0.5934816002845764, "learning_rate": 4.919195169092472e-06, "loss": 0.0187, "step": 69180 }, { "grad_norm": 0.15991389751434326, "learning_rate": 4.910259181179994e-06, "loss": 0.0138, "step": 69190 }, { "grad_norm": 0.238319531083107, "learning_rate": 4.901330897871015e-06, "loss": 0.0131, "step": 69200 }, { "grad_norm": 0.2728886604309082, "learning_rate": 4.8924103206911375e-06, "loss": 0.0111, "step": 69210 }, { "grad_norm": 0.23963294923305511, "learning_rate": 4.883497451164637e-06, "loss": 0.0216, "step": 69220 }, { "grad_norm": 0.18084914982318878, "learning_rate": 4.874592290814484e-06, "loss": 0.0115, "step": 69230 }, { "grad_norm": 0.6223146915435791, "learning_rate": 4.865694841162327e-06, "loss": 0.0259, "step": 69240 }, { "grad_norm": 0.19631460309028625, "learning_rate": 4.856805103728496e-06, "loss": 0.017, "step": 69250 }, { "grad_norm": 0.11840114742517471, "learning_rate": 4.847923080032007e-06, "loss": 0.0092, "step": 69260 }, { "grad_norm": 1.1382830142974854, "learning_rate": 4.8390487715905486e-06, "loss": 0.0107, "step": 69270 }, { "grad_norm": 0.7976599335670471, "learning_rate": 4.8301821799205e-06, "loss": 0.0094, "step": 69280 }, { "grad_norm": 0.24256205558776855, "learning_rate": 4.821323306536918e-06, "loss": 0.0112, "step": 69290 }, { "grad_norm": 0.16314797103405, "learning_rate": 4.8124721529535455e-06, "loss": 0.0104, "step": 69300 }, { "grad_norm": 0.23967160284519196, "learning_rate": 4.803628720682807e-06, "loss": 0.0152, "step": 69310 }, { "grad_norm": 0.26336413621902466, "learning_rate": 4.794793011235776e-06, "loss": 0.0141, "step": 69320 }, { "grad_norm": 0.18625520169734955, "learning_rate": 4.78596502612228e-06, "loss": 0.016, "step": 69330 }, { "grad_norm": 0.27828818559646606, "learning_rate": 4.777144766850738e-06, "loss": 0.0147, "step": 69340 }, { "grad_norm": 0.18093477189540863, "learning_rate": 4.768332234928313e-06, "loss": 0.014, "step": 69350 }, { "grad_norm": 0.08721614629030228, "learning_rate": 4.759527431860828e-06, "loss": 0.0118, "step": 69360 }, { "grad_norm": 0.24108350276947021, "learning_rate": 4.750730359152755e-06, "loss": 0.0151, "step": 69370 }, { "grad_norm": 0.14022643864154816, "learning_rate": 4.741941018307311e-06, "loss": 0.0128, "step": 69380 }, { "grad_norm": 0.17285212874412537, "learning_rate": 4.733159410826321e-06, "loss": 0.0168, "step": 69390 }, { "grad_norm": 0.2299620658159256, "learning_rate": 4.724385538210357e-06, "loss": 0.0118, "step": 69400 }, { "grad_norm": 0.3072717487812042, "learning_rate": 4.715619401958599e-06, "loss": 0.0149, "step": 69410 }, { "grad_norm": 0.9371174573898315, "learning_rate": 4.706861003568958e-06, "loss": 0.0162, "step": 69420 }, { "grad_norm": 0.34132495522499084, "learning_rate": 4.698110344538003e-06, "loss": 0.0131, "step": 69430 }, { "grad_norm": 0.25748124718666077, "learning_rate": 4.689367426360975e-06, "loss": 0.0126, "step": 69440 }, { "grad_norm": 0.30287355184555054, "learning_rate": 4.680632250531819e-06, "loss": 0.0146, "step": 69450 }, { "grad_norm": 0.21404802799224854, "learning_rate": 4.671904818543115e-06, "loss": 0.0128, "step": 69460 }, { "grad_norm": 0.30772557854652405, "learning_rate": 4.66318513188615e-06, "loss": 0.0124, "step": 69470 }, { "grad_norm": 0.16438987851142883, "learning_rate": 4.654473192050884e-06, "loss": 0.0115, "step": 69480 }, { "grad_norm": 0.2378910779953003, "learning_rate": 4.64576900052594e-06, "loss": 0.0096, "step": 69490 }, { "grad_norm": 0.23833870887756348, "learning_rate": 4.637072558798638e-06, "loss": 0.0108, "step": 69500 }, { "grad_norm": 0.19416993856430054, "learning_rate": 4.628383868354969e-06, "loss": 0.017, "step": 69510 }, { "grad_norm": 0.24971680343151093, "learning_rate": 4.6197029306795595e-06, "loss": 0.0081, "step": 69520 }, { "grad_norm": 0.37060755491256714, "learning_rate": 4.611029747255779e-06, "loss": 0.0088, "step": 69530 }, { "grad_norm": 0.2315511554479599, "learning_rate": 4.6023643195656164e-06, "loss": 0.0128, "step": 69540 }, { "grad_norm": 0.18545092642307281, "learning_rate": 4.593706649089768e-06, "loss": 0.0105, "step": 69550 }, { "grad_norm": 0.43332499265670776, "learning_rate": 4.585056737307597e-06, "loss": 0.0099, "step": 69560 }, { "grad_norm": 0.17732107639312744, "learning_rate": 4.576414585697103e-06, "loss": 0.0109, "step": 69570 }, { "grad_norm": 0.3469928801059723, "learning_rate": 4.567780195735044e-06, "loss": 0.015, "step": 69580 }, { "grad_norm": 0.21469178795814514, "learning_rate": 4.559153568896757e-06, "loss": 0.0124, "step": 69590 }, { "grad_norm": 0.34332478046417236, "learning_rate": 4.550534706656329e-06, "loss": 0.0144, "step": 69600 }, { "grad_norm": 0.29355481266975403, "learning_rate": 4.541923610486465e-06, "loss": 0.0111, "step": 69610 }, { "grad_norm": 0.2070053666830063, "learning_rate": 4.533320281858578e-06, "loss": 0.0129, "step": 69620 }, { "grad_norm": 0.2864062190055847, "learning_rate": 4.52472472224274e-06, "loss": 0.0161, "step": 69630 }, { "grad_norm": 0.22072958946228027, "learning_rate": 4.51613693310769e-06, "loss": 0.0112, "step": 69640 }, { "grad_norm": 0.2666764259338379, "learning_rate": 4.507556915920868e-06, "loss": 0.0118, "step": 69650 }, { "grad_norm": 0.31489893794059753, "learning_rate": 4.498984672148332e-06, "loss": 0.011, "step": 69660 }, { "grad_norm": 0.2610240876674652, "learning_rate": 4.490420203254864e-06, "loss": 0.0157, "step": 69670 }, { "grad_norm": 0.6307827234268188, "learning_rate": 4.481863510703893e-06, "loss": 0.019, "step": 69680 }, { "grad_norm": 0.19220374524593353, "learning_rate": 4.473314595957523e-06, "loss": 0.0121, "step": 69690 }, { "grad_norm": 0.2966868281364441, "learning_rate": 4.464773460476535e-06, "loss": 0.0125, "step": 69700 }, { "grad_norm": 0.19743812084197998, "learning_rate": 4.456240105720372e-06, "loss": 0.0111, "step": 69710 }, { "grad_norm": 0.19891998171806335, "learning_rate": 4.4477145331471405e-06, "loss": 0.0179, "step": 69720 }, { "grad_norm": 0.2100442796945572, "learning_rate": 4.439196744213653e-06, "loss": 0.0099, "step": 69730 }, { "grad_norm": 0.16325144469738007, "learning_rate": 4.430686740375339e-06, "loss": 0.0131, "step": 69740 }, { "grad_norm": 0.19026920199394226, "learning_rate": 4.422184523086342e-06, "loss": 0.01, "step": 69750 }, { "grad_norm": 0.25510692596435547, "learning_rate": 4.41369009379946e-06, "loss": 0.0105, "step": 69760 }, { "grad_norm": 0.18363933265209198, "learning_rate": 4.405203453966139e-06, "loss": 0.0147, "step": 69770 }, { "grad_norm": 0.428069144487381, "learning_rate": 4.396724605036539e-06, "loss": 0.0121, "step": 69780 }, { "grad_norm": 0.24710364639759064, "learning_rate": 4.388253548459437e-06, "loss": 0.0109, "step": 69790 }, { "grad_norm": 0.31595754623413086, "learning_rate": 4.3797902856823395e-06, "loss": 0.0123, "step": 69800 }, { "grad_norm": 0.15157485008239746, "learning_rate": 4.371334818151357e-06, "loss": 0.0095, "step": 69810 }, { "grad_norm": 0.32220345735549927, "learning_rate": 4.362887147311306e-06, "loss": 0.0132, "step": 69820 }, { "grad_norm": 0.22138632833957672, "learning_rate": 4.354447274605672e-06, "loss": 0.0119, "step": 69830 }, { "grad_norm": 0.15181398391723633, "learning_rate": 4.346015201476572e-06, "loss": 0.0203, "step": 69840 }, { "grad_norm": 0.22168366611003876, "learning_rate": 4.337590929364855e-06, "loss": 0.0105, "step": 69850 }, { "grad_norm": 0.24863062798976898, "learning_rate": 4.329174459709973e-06, "loss": 0.011, "step": 69860 }, { "grad_norm": 0.1640349179506302, "learning_rate": 4.320765793950071e-06, "loss": 0.0131, "step": 69870 }, { "grad_norm": 0.1980314552783966, "learning_rate": 4.312364933521962e-06, "loss": 0.0139, "step": 69880 }, { "grad_norm": 0.19017162919044495, "learning_rate": 4.303971879861129e-06, "loss": 0.012, "step": 69890 }, { "grad_norm": 0.38909339904785156, "learning_rate": 4.295586634401716e-06, "loss": 0.0155, "step": 69900 }, { "grad_norm": 0.22794581949710846, "learning_rate": 4.287209198576536e-06, "loss": 0.0097, "step": 69910 }, { "grad_norm": 0.16014301776885986, "learning_rate": 4.278839573817045e-06, "loss": 0.0149, "step": 69920 }, { "grad_norm": 0.3715759813785553, "learning_rate": 4.270477761553399e-06, "loss": 0.0132, "step": 69930 }, { "grad_norm": 0.2668314278125763, "learning_rate": 4.262123763214393e-06, "loss": 0.0094, "step": 69940 }, { "grad_norm": 0.2186616212129593, "learning_rate": 4.253777580227508e-06, "loss": 0.0101, "step": 69950 }, { "grad_norm": 0.45835933089256287, "learning_rate": 4.2454392140188755e-06, "loss": 0.0135, "step": 69960 }, { "grad_norm": 0.23008793592453003, "learning_rate": 4.2371086660132785e-06, "loss": 0.0125, "step": 69970 }, { "grad_norm": 0.25907468795776367, "learning_rate": 4.228785937634205e-06, "loss": 0.0101, "step": 69980 }, { "grad_norm": 0.21757841110229492, "learning_rate": 4.220471030303758e-06, "loss": 0.0106, "step": 69990 }, { "grad_norm": 0.24826708436012268, "learning_rate": 4.212163945442754e-06, "loss": 0.01, "step": 70000 }, { "grad_norm": 0.43954768776893616, "learning_rate": 4.203864684470621e-06, "loss": 0.0093, "step": 70010 }, { "grad_norm": 0.34544268250465393, "learning_rate": 4.19557324880549e-06, "loss": 0.0154, "step": 70020 }, { "grad_norm": 0.141774982213974, "learning_rate": 4.187289639864145e-06, "loss": 0.012, "step": 70030 }, { "grad_norm": 0.1977706402540207, "learning_rate": 4.1790138590619974e-06, "loss": 0.0107, "step": 70040 }, { "grad_norm": 0.3019197881221771, "learning_rate": 4.170745907813195e-06, "loss": 0.0157, "step": 70050 }, { "grad_norm": 0.21503795683383942, "learning_rate": 4.162485787530479e-06, "loss": 0.0128, "step": 70060 }, { "grad_norm": 0.41458773612976074, "learning_rate": 4.154233499625282e-06, "loss": 0.0132, "step": 70070 }, { "grad_norm": 0.1873917430639267, "learning_rate": 4.145989045507692e-06, "loss": 0.0163, "step": 70080 }, { "grad_norm": 0.6767328381538391, "learning_rate": 4.1377524265864666e-06, "loss": 0.0117, "step": 70090 }, { "grad_norm": 0.20334698259830475, "learning_rate": 4.1295236442690175e-06, "loss": 0.0105, "step": 70100 }, { "grad_norm": 0.37525975704193115, "learning_rate": 4.121302699961421e-06, "loss": 0.0115, "step": 70110 }, { "grad_norm": 0.302070677280426, "learning_rate": 4.113089595068403e-06, "loss": 0.0131, "step": 70120 }, { "grad_norm": 0.24493734538555145, "learning_rate": 4.104884330993364e-06, "loss": 0.0138, "step": 70130 }, { "grad_norm": 0.257058709859848, "learning_rate": 4.0966869091383585e-06, "loss": 0.0177, "step": 70140 }, { "grad_norm": 0.36905527114868164, "learning_rate": 4.088497330904101e-06, "loss": 0.0139, "step": 70150 }, { "grad_norm": 0.27273041009902954, "learning_rate": 4.080315597689976e-06, "loss": 0.0141, "step": 70160 }, { "grad_norm": 0.18713417649269104, "learning_rate": 4.072141710893995e-06, "loss": 0.0162, "step": 70170 }, { "grad_norm": 0.17992118000984192, "learning_rate": 4.063975671912879e-06, "loss": 0.0172, "step": 70180 }, { "grad_norm": 0.1829710602760315, "learning_rate": 4.055817482141949e-06, "loss": 0.0088, "step": 70190 }, { "grad_norm": 0.31374526023864746, "learning_rate": 4.047667142975259e-06, "loss": 0.0113, "step": 70200 }, { "grad_norm": 0.24353590607643127, "learning_rate": 4.039524655805443e-06, "loss": 0.0203, "step": 70210 }, { "grad_norm": 0.2373172640800476, "learning_rate": 4.03139002202384e-06, "loss": 0.0107, "step": 70220 }, { "grad_norm": 0.48417147994041443, "learning_rate": 4.023263243020447e-06, "loss": 0.0123, "step": 70230 }, { "grad_norm": 0.38489440083503723, "learning_rate": 4.015144320183884e-06, "loss": 0.0156, "step": 70240 }, { "grad_norm": 0.2500157356262207, "learning_rate": 4.007033254901482e-06, "loss": 0.0115, "step": 70250 }, { "grad_norm": 0.3019169867038727, "learning_rate": 3.9989300485591795e-06, "loss": 0.0128, "step": 70260 }, { "grad_norm": 0.1701507717370987, "learning_rate": 3.990834702541601e-06, "loss": 0.016, "step": 70270 }, { "grad_norm": 0.2933547794818878, "learning_rate": 3.982747218232019e-06, "loss": 0.0128, "step": 70280 }, { "grad_norm": 0.1783970296382904, "learning_rate": 3.97466759701236e-06, "loss": 0.0114, "step": 70290 }, { "grad_norm": 0.43826770782470703, "learning_rate": 3.966595840263226e-06, "loss": 0.0135, "step": 70300 }, { "grad_norm": 0.4085521399974823, "learning_rate": 3.958531949363836e-06, "loss": 0.0113, "step": 70310 }, { "grad_norm": 0.19255368411540985, "learning_rate": 3.950475925692098e-06, "loss": 0.0122, "step": 70320 }, { "grad_norm": 0.19078797101974487, "learning_rate": 3.9424277706245685e-06, "loss": 0.0128, "step": 70330 }, { "grad_norm": 0.17151156067848206, "learning_rate": 3.934387485536451e-06, "loss": 0.0105, "step": 70340 }, { "grad_norm": 0.44961124658584595, "learning_rate": 3.926355071801619e-06, "loss": 0.0099, "step": 70350 }, { "grad_norm": 0.2300688922405243, "learning_rate": 3.9183305307925965e-06, "loss": 0.0118, "step": 70360 }, { "grad_norm": 0.18892422318458557, "learning_rate": 3.910313863880533e-06, "loss": 0.0077, "step": 70370 }, { "grad_norm": 0.301209032535553, "learning_rate": 3.902305072435292e-06, "loss": 0.0112, "step": 70380 }, { "grad_norm": 0.32346588373184204, "learning_rate": 3.894304157825329e-06, "loss": 0.0112, "step": 70390 }, { "grad_norm": 0.28711560368537903, "learning_rate": 3.886311121417791e-06, "loss": 0.0131, "step": 70400 }, { "grad_norm": 0.1948072910308838, "learning_rate": 3.878325964578472e-06, "loss": 0.0103, "step": 70410 }, { "grad_norm": 0.190100759267807, "learning_rate": 3.870348688671815e-06, "loss": 0.0087, "step": 70420 }, { "grad_norm": 0.2565139830112457, "learning_rate": 3.862379295060931e-06, "loss": 0.0106, "step": 70430 }, { "grad_norm": 0.36875104904174805, "learning_rate": 3.854417785107534e-06, "loss": 0.0218, "step": 70440 }, { "grad_norm": 0.3140805661678314, "learning_rate": 3.8464641601720755e-06, "loss": 0.0109, "step": 70450 }, { "grad_norm": 0.24594655632972717, "learning_rate": 3.838518421613579e-06, "loss": 0.0141, "step": 70460 }, { "grad_norm": 0.5865241885185242, "learning_rate": 3.830580570789766e-06, "loss": 0.017, "step": 70470 }, { "grad_norm": 0.30035877227783203, "learning_rate": 3.822650609057005e-06, "loss": 0.01, "step": 70480 }, { "grad_norm": 0.3410014808177948, "learning_rate": 3.814728537770285e-06, "loss": 0.0122, "step": 70490 }, { "grad_norm": 0.9824566841125488, "learning_rate": 3.806814358283306e-06, "loss": 0.0142, "step": 70500 }, { "grad_norm": 0.4207300841808319, "learning_rate": 3.7989080719483596e-06, "loss": 0.0121, "step": 70510 }, { "grad_norm": 0.10794796794652939, "learning_rate": 3.7910096801164143e-06, "loss": 0.0163, "step": 70520 }, { "grad_norm": 0.17988404631614685, "learning_rate": 3.7831191841371016e-06, "loss": 0.0108, "step": 70530 }, { "grad_norm": 0.26076021790504456, "learning_rate": 3.775236585358688e-06, "loss": 0.0136, "step": 70540 }, { "grad_norm": 0.10544079542160034, "learning_rate": 3.7673618851280843e-06, "loss": 0.0131, "step": 70550 }, { "grad_norm": 0.4044678211212158, "learning_rate": 3.759495084790887e-06, "loss": 0.0114, "step": 70560 }, { "grad_norm": 0.2168959081172943, "learning_rate": 3.751636185691282e-06, "loss": 0.0143, "step": 70570 }, { "grad_norm": 0.33906713128089905, "learning_rate": 3.7437851891721607e-06, "loss": 0.0135, "step": 70580 }, { "grad_norm": 0.5255993008613586, "learning_rate": 3.7359420965750404e-06, "loss": 0.0131, "step": 70590 }, { "grad_norm": 0.18924637138843536, "learning_rate": 3.7281069092400922e-06, "loss": 0.0115, "step": 70600 }, { "grad_norm": 0.502548098564148, "learning_rate": 3.7202796285061348e-06, "loss": 0.0125, "step": 70610 }, { "grad_norm": 0.274368017911911, "learning_rate": 3.712460255710637e-06, "loss": 0.0144, "step": 70620 }, { "grad_norm": 0.1613813042640686, "learning_rate": 3.704648792189719e-06, "loss": 0.0093, "step": 70630 }, { "grad_norm": 0.5521836280822754, "learning_rate": 3.696845239278124e-06, "loss": 0.0121, "step": 70640 }, { "grad_norm": 0.2817066013813019, "learning_rate": 3.689049598309302e-06, "loss": 0.0142, "step": 70650 }, { "grad_norm": 0.2259397804737091, "learning_rate": 3.681261870615288e-06, "loss": 0.021, "step": 70660 }, { "grad_norm": 0.2434130758047104, "learning_rate": 3.6734820575268004e-06, "loss": 0.0216, "step": 70670 }, { "grad_norm": 0.11453516781330109, "learning_rate": 3.665710160373209e-06, "loss": 0.0109, "step": 70680 }, { "grad_norm": 0.27360084652900696, "learning_rate": 3.65794618048248e-06, "loss": 0.0104, "step": 70690 }, { "grad_norm": 0.18555279076099396, "learning_rate": 3.6501901191813125e-06, "loss": 0.01, "step": 70700 }, { "grad_norm": 0.15573734045028687, "learning_rate": 3.642441977794975e-06, "loss": 0.0145, "step": 70710 }, { "grad_norm": 0.2564886212348938, "learning_rate": 3.634701757647424e-06, "loss": 0.0152, "step": 70720 }, { "grad_norm": 0.4455163776874542, "learning_rate": 3.6269694600612468e-06, "loss": 0.0108, "step": 70730 }, { "grad_norm": 0.2579069137573242, "learning_rate": 3.619245086357681e-06, "loss": 0.0116, "step": 70740 }, { "grad_norm": 0.2588376998901367, "learning_rate": 3.611528637856615e-06, "loss": 0.0136, "step": 70750 }, { "grad_norm": 0.6632508635520935, "learning_rate": 3.6038201158765884e-06, "loss": 0.0109, "step": 70760 }, { "grad_norm": 0.515275239944458, "learning_rate": 3.5961195217347534e-06, "loss": 0.0198, "step": 70770 }, { "grad_norm": 0.17248718440532684, "learning_rate": 3.588426856746946e-06, "loss": 0.011, "step": 70780 }, { "grad_norm": 0.19559991359710693, "learning_rate": 3.5807421222276316e-06, "loss": 0.0171, "step": 70790 }, { "grad_norm": 0.2612893581390381, "learning_rate": 3.5730653194899156e-06, "loss": 0.0147, "step": 70800 }, { "grad_norm": 0.1185828298330307, "learning_rate": 3.565396449845554e-06, "loss": 0.008, "step": 70810 }, { "grad_norm": 0.30640995502471924, "learning_rate": 3.557735514604954e-06, "loss": 0.0126, "step": 70820 }, { "grad_norm": 0.4705408811569214, "learning_rate": 3.5500825150771633e-06, "loss": 0.0126, "step": 70830 }, { "grad_norm": 0.1335756629705429, "learning_rate": 3.5424374525698466e-06, "loss": 0.0126, "step": 70840 }, { "grad_norm": 0.2731741964817047, "learning_rate": 3.5348003283893704e-06, "loss": 0.012, "step": 70850 }, { "grad_norm": 0.19061189889907837, "learning_rate": 3.527171143840685e-06, "loss": 0.0127, "step": 70860 }, { "grad_norm": 0.16216368973255157, "learning_rate": 3.519549900227409e-06, "loss": 0.0137, "step": 70870 }, { "grad_norm": 0.23558661341667175, "learning_rate": 3.511936598851828e-06, "loss": 0.0104, "step": 70880 }, { "grad_norm": 0.14590276777744293, "learning_rate": 3.504331241014813e-06, "loss": 0.0134, "step": 70890 }, { "grad_norm": 0.5023281574249268, "learning_rate": 3.4967338280159414e-06, "loss": 0.0115, "step": 70900 }, { "grad_norm": 0.2723047733306885, "learning_rate": 3.4891443611533846e-06, "loss": 0.016, "step": 70910 }, { "grad_norm": 0.20318561792373657, "learning_rate": 3.481562841723984e-06, "loss": 0.014, "step": 70920 }, { "grad_norm": 0.31887561082839966, "learning_rate": 3.473989271023215e-06, "loss": 0.0112, "step": 70930 }, { "grad_norm": 0.5176720023155212, "learning_rate": 3.4664236503451854e-06, "loss": 0.0162, "step": 70940 }, { "grad_norm": 0.16077543795108795, "learning_rate": 3.4588659809826674e-06, "loss": 0.008, "step": 70950 }, { "grad_norm": 0.24809981882572174, "learning_rate": 3.451316264227039e-06, "loss": 0.0161, "step": 70960 }, { "grad_norm": 0.24561172723770142, "learning_rate": 3.4437745013683574e-06, "loss": 0.0102, "step": 70970 }, { "grad_norm": 0.21793866157531738, "learning_rate": 3.4362406936952916e-06, "loss": 0.0117, "step": 70980 }, { "grad_norm": 0.49490591883659363, "learning_rate": 3.4287148424951786e-06, "loss": 0.0113, "step": 70990 }, { "grad_norm": 0.156906858086586, "learning_rate": 3.4211969490539617e-06, "loss": 0.012, "step": 71000 }, { "grad_norm": 0.3045717775821686, "learning_rate": 3.4136870146562584e-06, "loss": 0.0131, "step": 71010 }, { "grad_norm": 0.3820546567440033, "learning_rate": 3.406185040585308e-06, "loss": 0.0121, "step": 71020 }, { "grad_norm": 0.20953799784183502, "learning_rate": 3.3986910281229966e-06, "loss": 0.0142, "step": 71030 }, { "grad_norm": 0.3535473346710205, "learning_rate": 3.391204978549828e-06, "loss": 0.0158, "step": 71040 }, { "grad_norm": 0.3474843502044678, "learning_rate": 3.3837268931449785e-06, "loss": 0.0181, "step": 71050 }, { "grad_norm": 0.2784714698791504, "learning_rate": 3.376256773186248e-06, "loss": 0.0143, "step": 71060 }, { "grad_norm": 0.1943686157464981, "learning_rate": 3.3687946199500664e-06, "loss": 0.0135, "step": 71070 }, { "grad_norm": 0.27036190032958984, "learning_rate": 3.3613404347115295e-06, "loss": 0.0106, "step": 71080 }, { "grad_norm": 0.24294303357601166, "learning_rate": 3.35389421874433e-06, "loss": 0.0104, "step": 71090 }, { "grad_norm": 0.1438937783241272, "learning_rate": 3.3464559733208446e-06, "loss": 0.0097, "step": 71100 }, { "grad_norm": 0.3246172368526459, "learning_rate": 3.3390256997120505e-06, "loss": 0.0104, "step": 71110 }, { "grad_norm": 0.23081564903259277, "learning_rate": 3.331603399187583e-06, "loss": 0.0113, "step": 71120 }, { "grad_norm": 0.1651192605495453, "learning_rate": 3.324189073015721e-06, "loss": 0.0096, "step": 71130 }, { "grad_norm": 0.16741856932640076, "learning_rate": 3.3167827224633465e-06, "loss": 0.0117, "step": 71140 }, { "grad_norm": 0.21294566988945007, "learning_rate": 3.309384348796024e-06, "loss": 0.0107, "step": 71150 }, { "grad_norm": 0.2292691171169281, "learning_rate": 3.301993953277921e-06, "loss": 0.0146, "step": 71160 }, { "grad_norm": 0.41984668374061584, "learning_rate": 3.29461153717186e-06, "loss": 0.0111, "step": 71170 }, { "grad_norm": 0.2793019115924835, "learning_rate": 3.287237101739293e-06, "loss": 0.0192, "step": 71180 }, { "grad_norm": 0.24137601256370544, "learning_rate": 3.2798706482403075e-06, "loss": 0.0141, "step": 71190 }, { "grad_norm": 0.2215675413608551, "learning_rate": 3.2725121779336285e-06, "loss": 0.0144, "step": 71200 }, { "grad_norm": 0.2750155031681061, "learning_rate": 3.265161692076618e-06, "loss": 0.0133, "step": 71210 }, { "grad_norm": 0.09107202291488647, "learning_rate": 3.257819191925282e-06, "loss": 0.0103, "step": 71220 }, { "grad_norm": 0.20635829865932465, "learning_rate": 3.2504846787342392e-06, "loss": 0.0124, "step": 71230 }, { "grad_norm": 0.1161121279001236, "learning_rate": 3.243158153756759e-06, "loss": 0.0128, "step": 71240 }, { "grad_norm": 0.2989995777606964, "learning_rate": 3.2358396182447514e-06, "loss": 0.0105, "step": 71250 }, { "grad_norm": 2.7478997707366943, "learning_rate": 3.2285290734487496e-06, "loss": 0.01, "step": 71260 }, { "grad_norm": 0.3621162176132202, "learning_rate": 3.221226520617926e-06, "loss": 0.0107, "step": 71270 }, { "grad_norm": 0.2164556086063385, "learning_rate": 3.213931961000094e-06, "loss": 0.0175, "step": 71280 }, { "grad_norm": 0.4094865918159485, "learning_rate": 3.2066453958416733e-06, "loss": 0.0116, "step": 71290 }, { "grad_norm": 0.2224550098180771, "learning_rate": 3.199366826387773e-06, "loss": 0.0173, "step": 71300 }, { "grad_norm": 0.21781396865844727, "learning_rate": 3.192096253882071e-06, "loss": 0.0157, "step": 71310 }, { "grad_norm": 0.31698113679885864, "learning_rate": 3.1848336795669177e-06, "loss": 0.0137, "step": 71320 }, { "grad_norm": 0.19778025150299072, "learning_rate": 3.1775791046833035e-06, "loss": 0.0119, "step": 71330 }, { "grad_norm": 0.18747267127037048, "learning_rate": 3.170332530470804e-06, "loss": 0.013, "step": 71340 }, { "grad_norm": 0.32808682322502136, "learning_rate": 3.1630939581677012e-06, "loss": 0.0163, "step": 71350 }, { "grad_norm": 0.2886841893196106, "learning_rate": 3.155863389010838e-06, "loss": 0.0082, "step": 71360 }, { "grad_norm": 0.7806932330131531, "learning_rate": 3.1486408242357323e-06, "loss": 0.0151, "step": 71370 }, { "grad_norm": 0.929550290107727, "learning_rate": 3.1414262650765248e-06, "loss": 0.0238, "step": 71380 }, { "grad_norm": 0.28877466917037964, "learning_rate": 3.134219712765979e-06, "loss": 0.0152, "step": 71390 }, { "grad_norm": 0.5541632771492004, "learning_rate": 3.1270211685355044e-06, "loss": 0.016, "step": 71400 }, { "grad_norm": 0.3629436790943146, "learning_rate": 3.1198306336151338e-06, "loss": 0.012, "step": 71410 }, { "grad_norm": 0.2474457323551178, "learning_rate": 3.11264810923354e-06, "loss": 0.013, "step": 71420 }, { "grad_norm": 0.16338181495666504, "learning_rate": 3.105473596618008e-06, "loss": 0.0104, "step": 71430 }, { "grad_norm": 0.2531082332134247, "learning_rate": 3.0983070969944683e-06, "loss": 0.0155, "step": 71440 }, { "grad_norm": 0.2245549112558365, "learning_rate": 3.09114861158748e-06, "loss": 0.01, "step": 71450 }, { "grad_norm": 0.2002027928829193, "learning_rate": 3.0839981416202314e-06, "loss": 0.0169, "step": 71460 }, { "grad_norm": 0.22668318450450897, "learning_rate": 3.0768556883145507e-06, "loss": 0.0119, "step": 71470 }, { "grad_norm": 0.3595426678657532, "learning_rate": 3.0697212528908834e-06, "loss": 0.0152, "step": 71480 }, { "grad_norm": 0.27862685918807983, "learning_rate": 3.0625948365683e-06, "loss": 0.0108, "step": 71490 }, { "grad_norm": 0.1305590122938156, "learning_rate": 3.055476440564531e-06, "loss": 0.0168, "step": 71500 }, { "grad_norm": 0.13166742026805878, "learning_rate": 3.0483660660958924e-06, "loss": 0.0121, "step": 71510 }, { "grad_norm": 0.34781527519226074, "learning_rate": 3.0412637143773624e-06, "loss": 0.0149, "step": 71520 }, { "grad_norm": 0.20552502572536469, "learning_rate": 3.034169386622554e-06, "loss": 0.0125, "step": 71530 }, { "grad_norm": 0.22518028318881989, "learning_rate": 3.027083084043658e-06, "loss": 0.0112, "step": 71540 }, { "grad_norm": 0.3123595714569092, "learning_rate": 3.0200048078515676e-06, "loss": 0.0143, "step": 71550 }, { "grad_norm": 0.22603517770767212, "learning_rate": 3.012934559255742e-06, "loss": 0.0111, "step": 71560 }, { "grad_norm": 0.24503996968269348, "learning_rate": 3.0058723394642994e-06, "loss": 0.0156, "step": 71570 }, { "grad_norm": 0.16713877022266388, "learning_rate": 2.998818149683985e-06, "loss": 0.015, "step": 71580 }, { "grad_norm": 0.2632477283477783, "learning_rate": 2.9917719911201627e-06, "loss": 0.012, "step": 71590 }, { "grad_norm": 0.4929274320602417, "learning_rate": 2.9847338649768352e-06, "loss": 0.0118, "step": 71600 }, { "grad_norm": 0.12690085172653198, "learning_rate": 2.977703772456608e-06, "loss": 0.0117, "step": 71610 }, { "grad_norm": 0.2250606119632721, "learning_rate": 2.9706817147607535e-06, "loss": 0.0112, "step": 71620 }, { "grad_norm": 0.2885238528251648, "learning_rate": 2.9636676930891393e-06, "loss": 0.0118, "step": 71630 }, { "grad_norm": 0.21055996417999268, "learning_rate": 2.9566617086402625e-06, "loss": 0.0091, "step": 71640 }, { "grad_norm": 0.3505319654941559, "learning_rate": 2.9496637626112655e-06, "loss": 0.0111, "step": 71650 }, { "grad_norm": 0.32050204277038574, "learning_rate": 2.9426738561979027e-06, "loss": 0.0099, "step": 71660 }, { "grad_norm": 0.3828418552875519, "learning_rate": 2.9356919905945524e-06, "loss": 0.027, "step": 71670 }, { "grad_norm": 0.21862289309501648, "learning_rate": 2.928718166994243e-06, "loss": 0.0112, "step": 71680 }, { "grad_norm": 0.22643399238586426, "learning_rate": 2.9217523865885833e-06, "loss": 0.0156, "step": 71690 }, { "grad_norm": 0.18647028505802155, "learning_rate": 2.9147946505678537e-06, "loss": 0.0118, "step": 71700 }, { "grad_norm": 0.24946007132530212, "learning_rate": 2.9078449601209313e-06, "loss": 0.0092, "step": 71710 }, { "grad_norm": 0.590698778629303, "learning_rate": 2.900903316435333e-06, "loss": 0.018, "step": 71720 }, { "grad_norm": 0.8687319755554199, "learning_rate": 2.8939697206971983e-06, "loss": 0.015, "step": 71730 }, { "grad_norm": 0.23174633085727692, "learning_rate": 2.8870441740912746e-06, "loss": 0.0121, "step": 71740 }, { "grad_norm": 0.3975253105163574, "learning_rate": 2.8801266778009762e-06, "loss": 0.0136, "step": 71750 }, { "grad_norm": 0.20948165655136108, "learning_rate": 2.8732172330082798e-06, "loss": 0.0146, "step": 71760 }, { "grad_norm": 0.23595650494098663, "learning_rate": 2.8663158408938517e-06, "loss": 0.0147, "step": 71770 }, { "grad_norm": 0.7165959477424622, "learning_rate": 2.859422502636938e-06, "loss": 0.0116, "step": 71780 }, { "grad_norm": 0.3003486394882202, "learning_rate": 2.8525372194154076e-06, "loss": 0.0122, "step": 71790 }, { "grad_norm": 0.1593303382396698, "learning_rate": 2.8456599924057913e-06, "loss": 0.0113, "step": 71800 }, { "grad_norm": 0.14144006371498108, "learning_rate": 2.8387908227831995e-06, "loss": 0.0113, "step": 71810 }, { "grad_norm": 0.18691091239452362, "learning_rate": 2.831929711721404e-06, "loss": 0.0125, "step": 71820 }, { "grad_norm": 0.19236907362937927, "learning_rate": 2.825076660392767e-06, "loss": 0.0113, "step": 71830 }, { "grad_norm": 0.4731936752796173, "learning_rate": 2.8182316699682908e-06, "loss": 0.0177, "step": 71840 }, { "grad_norm": 0.4447687268257141, "learning_rate": 2.811394741617601e-06, "loss": 0.0149, "step": 71850 }, { "grad_norm": 0.43386149406433105, "learning_rate": 2.8045658765089356e-06, "loss": 0.0146, "step": 71860 }, { "grad_norm": 0.1912773847579956, "learning_rate": 2.7977450758091605e-06, "loss": 0.0179, "step": 71870 }, { "grad_norm": 0.43049323558807373, "learning_rate": 2.790932340683783e-06, "loss": 0.0153, "step": 71880 }, { "grad_norm": 0.24432028830051422, "learning_rate": 2.7841276722968823e-06, "loss": 0.0154, "step": 71890 }, { "grad_norm": 0.3693300187587738, "learning_rate": 2.7773310718112067e-06, "loss": 0.0108, "step": 71900 }, { "grad_norm": 0.20832601189613342, "learning_rate": 2.7705425403881102e-06, "loss": 0.0143, "step": 71910 }, { "grad_norm": 0.19189585745334625, "learning_rate": 2.763762079187565e-06, "loss": 0.0107, "step": 71920 }, { "grad_norm": 0.2832629680633545, "learning_rate": 2.756989689368178e-06, "loss": 0.0136, "step": 71930 }, { "grad_norm": 0.17114122211933136, "learning_rate": 2.750225372087134e-06, "loss": 0.0135, "step": 71940 }, { "grad_norm": 0.3194054961204529, "learning_rate": 2.7434691285003033e-06, "loss": 0.0125, "step": 71950 }, { "grad_norm": 0.33004486560821533, "learning_rate": 2.7367209597621178e-06, "loss": 0.013, "step": 71960 }, { "grad_norm": 0.23412884771823883, "learning_rate": 2.729980867025683e-06, "loss": 0.0171, "step": 71970 }, { "grad_norm": 0.24210339784622192, "learning_rate": 2.7232488514426724e-06, "loss": 0.0141, "step": 71980 }, { "grad_norm": 0.15535464882850647, "learning_rate": 2.7165249141634097e-06, "loss": 0.0112, "step": 71990 }, { "grad_norm": 0.4560002386569977, "learning_rate": 2.709809056336837e-06, "loss": 0.0145, "step": 72000 }, { "grad_norm": 0.2173299491405487, "learning_rate": 2.703101279110498e-06, "loss": 0.0112, "step": 72010 }, { "grad_norm": 0.1924637407064438, "learning_rate": 2.696401583630587e-06, "loss": 0.0134, "step": 72020 }, { "grad_norm": 0.35286423563957214, "learning_rate": 2.689709971041887e-06, "loss": 0.012, "step": 72030 }, { "grad_norm": 0.4314091205596924, "learning_rate": 2.6830264424878126e-06, "loss": 0.0141, "step": 72040 }, { "grad_norm": 0.29806452989578247, "learning_rate": 2.676350999110394e-06, "loss": 0.0124, "step": 72050 }, { "grad_norm": 0.15993335843086243, "learning_rate": 2.6696836420502856e-06, "loss": 0.0198, "step": 72060 }, { "grad_norm": 0.17528532445430756, "learning_rate": 2.663024372446765e-06, "loss": 0.0103, "step": 72070 }, { "grad_norm": 0.23171919584274292, "learning_rate": 2.6563731914377055e-06, "loss": 0.0102, "step": 72080 }, { "grad_norm": 0.21983499825000763, "learning_rate": 2.6497301001596087e-06, "loss": 0.022, "step": 72090 }, { "grad_norm": 0.46691620349884033, "learning_rate": 2.643095099747611e-06, "loss": 0.0185, "step": 72100 }, { "grad_norm": 0.26186802983283997, "learning_rate": 2.636468191335445e-06, "loss": 0.0133, "step": 72110 }, { "grad_norm": 0.20431740581989288, "learning_rate": 2.629849376055471e-06, "loss": 0.0148, "step": 72120 }, { "grad_norm": 0.14623461663722992, "learning_rate": 2.6232386550386678e-06, "loss": 0.0119, "step": 72130 }, { "grad_norm": 0.34354788064956665, "learning_rate": 2.6166360294146097e-06, "loss": 0.0107, "step": 72140 }, { "grad_norm": 0.4252845346927643, "learning_rate": 2.6100415003115275e-06, "loss": 0.0184, "step": 72150 }, { "grad_norm": 0.38187164068222046, "learning_rate": 2.603455068856225e-06, "loss": 0.0145, "step": 72160 }, { "grad_norm": 0.22004979848861694, "learning_rate": 2.5968767361741584e-06, "loss": 0.0136, "step": 72170 }, { "grad_norm": 0.27051010727882385, "learning_rate": 2.590306503389378e-06, "loss": 0.0146, "step": 72180 }, { "grad_norm": 0.14611172676086426, "learning_rate": 2.583744371624558e-06, "loss": 0.0212, "step": 72190 }, { "grad_norm": 0.24723941087722778, "learning_rate": 2.57719034200099e-06, "loss": 0.0117, "step": 72200 }, { "grad_norm": 0.1794895976781845, "learning_rate": 2.570644415638568e-06, "loss": 0.0137, "step": 72210 }, { "grad_norm": 0.19567111134529114, "learning_rate": 2.5641065936558295e-06, "loss": 0.0164, "step": 72220 }, { "grad_norm": 0.3486214578151703, "learning_rate": 2.557576877169898e-06, "loss": 0.014, "step": 72230 }, { "grad_norm": 0.12937764823436737, "learning_rate": 2.5510552672965205e-06, "loss": 0.0074, "step": 72240 }, { "grad_norm": 0.3748423457145691, "learning_rate": 2.544541765150077e-06, "loss": 0.0113, "step": 72250 }, { "grad_norm": 0.2550032436847687, "learning_rate": 2.5380363718435165e-06, "loss": 0.0142, "step": 72260 }, { "grad_norm": 0.38760077953338623, "learning_rate": 2.5315390884884714e-06, "loss": 0.0132, "step": 72270 }, { "grad_norm": 0.1716487556695938, "learning_rate": 2.5250499161951214e-06, "loss": 0.0123, "step": 72280 }, { "grad_norm": 0.2630889415740967, "learning_rate": 2.518568856072301e-06, "loss": 0.0107, "step": 72290 }, { "grad_norm": 0.1653032898902893, "learning_rate": 2.512095909227441e-06, "loss": 0.0124, "step": 72300 }, { "grad_norm": 0.2056257128715515, "learning_rate": 2.50563107676659e-06, "loss": 0.0152, "step": 72310 }, { "grad_norm": 0.22635485231876373, "learning_rate": 2.4991743597944152e-06, "loss": 0.0124, "step": 72320 }, { "grad_norm": 0.10660237818956375, "learning_rate": 2.492725759414205e-06, "loss": 0.0126, "step": 72330 }, { "grad_norm": 0.26397213339805603, "learning_rate": 2.4862852767278234e-06, "loss": 0.0106, "step": 72340 }, { "grad_norm": 0.10999469459056854, "learning_rate": 2.479852912835784e-06, "loss": 0.0086, "step": 72350 }, { "grad_norm": 0.29271960258483887, "learning_rate": 2.4734286688372075e-06, "loss": 0.0205, "step": 72360 }, { "grad_norm": 0.3200758397579193, "learning_rate": 2.467012545829811e-06, "loss": 0.0159, "step": 72370 }, { "grad_norm": 0.26067712903022766, "learning_rate": 2.460604544909945e-06, "loss": 0.0104, "step": 72380 }, { "grad_norm": 0.2690565586090088, "learning_rate": 2.454204667172555e-06, "loss": 0.0085, "step": 72390 }, { "grad_norm": 0.239073246717453, "learning_rate": 2.447812913711217e-06, "loss": 0.0134, "step": 72400 }, { "grad_norm": 0.18661439418792725, "learning_rate": 2.44142928561808e-06, "loss": 0.0127, "step": 72410 }, { "grad_norm": 0.1980850249528885, "learning_rate": 2.435053783983965e-06, "loss": 0.0098, "step": 72420 }, { "grad_norm": 0.3772798776626587, "learning_rate": 2.4286864098982453e-06, "loss": 0.0163, "step": 72430 }, { "grad_norm": 0.4759204089641571, "learning_rate": 2.4223271644489397e-06, "loss": 0.0118, "step": 72440 }, { "grad_norm": 0.13619904220104218, "learning_rate": 2.415976048722679e-06, "loss": 0.0074, "step": 72450 }, { "grad_norm": 0.3688609302043915, "learning_rate": 2.4096330638046673e-06, "loss": 0.014, "step": 72460 }, { "grad_norm": 0.2701166272163391, "learning_rate": 2.4032982107787816e-06, "loss": 0.013, "step": 72470 }, { "grad_norm": 0.19186483323574066, "learning_rate": 2.396971490727451e-06, "loss": 0.0123, "step": 72480 }, { "grad_norm": 0.2749800980091095, "learning_rate": 2.3906529047317493e-06, "loss": 0.0094, "step": 72490 }, { "grad_norm": 0.3102211356163025, "learning_rate": 2.3843424538713465e-06, "loss": 0.0117, "step": 72500 }, { "grad_norm": 0.3769700527191162, "learning_rate": 2.3780401392245298e-06, "loss": 0.0143, "step": 72510 }, { "grad_norm": 0.361640989780426, "learning_rate": 2.3717459618681883e-06, "loss": 0.0126, "step": 72520 }, { "grad_norm": 0.19804121553897858, "learning_rate": 2.3654599228778274e-06, "loss": 0.014, "step": 72530 }, { "grad_norm": 0.21681615710258484, "learning_rate": 2.3591820233275607e-06, "loss": 0.0133, "step": 72540 }, { "grad_norm": 0.374228835105896, "learning_rate": 2.3529122642901024e-06, "loss": 0.0158, "step": 72550 }, { "grad_norm": 0.32426324486732483, "learning_rate": 2.3466506468367845e-06, "loss": 0.0226, "step": 72560 }, { "grad_norm": 0.2968634068965912, "learning_rate": 2.3403971720375506e-06, "loss": 0.0122, "step": 72570 }, { "grad_norm": 0.13106967508792877, "learning_rate": 2.334151840960952e-06, "loss": 0.0079, "step": 72580 }, { "grad_norm": 0.25464409589767456, "learning_rate": 2.3279146546741347e-06, "loss": 0.0108, "step": 72590 }, { "grad_norm": 0.3795416057109833, "learning_rate": 2.32168561424288e-06, "loss": 0.0121, "step": 72600 }, { "grad_norm": 0.1575748175382614, "learning_rate": 2.3154647207315307e-06, "loss": 0.0103, "step": 72610 }, { "grad_norm": 0.1212276965379715, "learning_rate": 2.309251975203103e-06, "loss": 0.0112, "step": 72620 }, { "grad_norm": 0.18014760315418243, "learning_rate": 2.303047378719159e-06, "loss": 0.0177, "step": 72630 }, { "grad_norm": 0.21940398216247559, "learning_rate": 2.2968509323399e-06, "loss": 0.0119, "step": 72640 }, { "grad_norm": 0.12469609081745148, "learning_rate": 2.290662637124147e-06, "loss": 0.0104, "step": 72650 }, { "grad_norm": 0.2129245400428772, "learning_rate": 2.2844824941292807e-06, "loss": 0.0088, "step": 72660 }, { "grad_norm": 0.275682270526886, "learning_rate": 2.2783105044113406e-06, "loss": 0.0141, "step": 72670 }, { "grad_norm": 0.27807971835136414, "learning_rate": 2.272146669024944e-06, "loss": 0.0096, "step": 72680 }, { "grad_norm": 0.1768648773431778, "learning_rate": 2.26599098902332e-06, "loss": 0.0105, "step": 72690 }, { "grad_norm": 0.16321487724781036, "learning_rate": 2.2598434654583113e-06, "loss": 0.0091, "step": 72700 }, { "grad_norm": 0.3257555663585663, "learning_rate": 2.253704099380355e-06, "loss": 0.0142, "step": 72710 }, { "grad_norm": 0.2514015734195709, "learning_rate": 2.247572891838512e-06, "loss": 0.0155, "step": 72720 }, { "grad_norm": 0.30712467432022095, "learning_rate": 2.241449843880422e-06, "loss": 0.0118, "step": 72730 }, { "grad_norm": 0.09237181395292282, "learning_rate": 2.235334956552354e-06, "loss": 0.017, "step": 72740 }, { "grad_norm": 0.18114158511161804, "learning_rate": 2.2292282308991775e-06, "loss": 0.0141, "step": 72750 }, { "grad_norm": 0.3750894367694855, "learning_rate": 2.223129667964363e-06, "loss": 0.0109, "step": 72760 }, { "grad_norm": 0.16107258200645447, "learning_rate": 2.2170392687899834e-06, "loss": 0.0093, "step": 72770 }, { "grad_norm": 0.28792881965637207, "learning_rate": 2.210957034416733e-06, "loss": 0.0095, "step": 72780 }, { "grad_norm": 0.05682104080915451, "learning_rate": 2.2048829658838867e-06, "loss": 0.0112, "step": 72790 }, { "grad_norm": 0.23273754119873047, "learning_rate": 2.1988170642293525e-06, "loss": 0.0104, "step": 72800 }, { "grad_norm": 0.11673016846179962, "learning_rate": 2.1927593304896075e-06, "loss": 0.0114, "step": 72810 }, { "grad_norm": 0.1739816814661026, "learning_rate": 2.1867097656997626e-06, "loss": 0.0117, "step": 72820 }, { "grad_norm": 0.211011603474617, "learning_rate": 2.180668370893524e-06, "loss": 0.0119, "step": 72830 }, { "grad_norm": 0.38384783267974854, "learning_rate": 2.174635147103199e-06, "loss": 0.0147, "step": 72840 }, { "grad_norm": 0.20810210704803467, "learning_rate": 2.1686100953597075e-06, "loss": 0.0089, "step": 72850 }, { "grad_norm": 0.30519866943359375, "learning_rate": 2.1625932166925433e-06, "loss": 0.0135, "step": 72860 }, { "grad_norm": 0.3027620017528534, "learning_rate": 2.1565845121298556e-06, "loss": 0.0135, "step": 72870 }, { "grad_norm": 0.17331482470035553, "learning_rate": 2.150583982698351e-06, "loss": 0.0124, "step": 72880 }, { "grad_norm": 0.2667638957500458, "learning_rate": 2.144591629423359e-06, "loss": 0.0089, "step": 72890 }, { "grad_norm": 0.24891887605190277, "learning_rate": 2.138607453328817e-06, "loss": 0.0146, "step": 72900 }, { "grad_norm": 0.156494602560997, "learning_rate": 2.132631455437234e-06, "loss": 0.0129, "step": 72910 }, { "grad_norm": 0.2682884633541107, "learning_rate": 2.1266636367697714e-06, "loss": 0.0154, "step": 72920 }, { "grad_norm": 0.35581162571907043, "learning_rate": 2.120703998346152e-06, "loss": 0.0103, "step": 72930 }, { "grad_norm": 0.1972971260547638, "learning_rate": 2.1147525411847114e-06, "loss": 0.0116, "step": 72940 }, { "grad_norm": 0.2422160506248474, "learning_rate": 2.108809266302403e-06, "loss": 0.0103, "step": 72950 }, { "grad_norm": 0.21443353593349457, "learning_rate": 2.102874174714764e-06, "loss": 0.0115, "step": 72960 }, { "grad_norm": 0.31243136525154114, "learning_rate": 2.0969472674359393e-06, "loss": 0.0143, "step": 72970 }, { "grad_norm": 0.2498828023672104, "learning_rate": 2.091028545478668e-06, "loss": 0.0156, "step": 72980 }, { "grad_norm": 0.20454901456832886, "learning_rate": 2.08511800985432e-06, "loss": 0.0084, "step": 72990 }, { "grad_norm": 0.25531378388404846, "learning_rate": 2.07921566157282e-06, "loss": 0.0183, "step": 73000 }, { "grad_norm": 0.26521509885787964, "learning_rate": 2.073321501642728e-06, "loss": 0.0105, "step": 73010 }, { "grad_norm": 0.20823156833648682, "learning_rate": 2.0674355310711937e-06, "loss": 0.0108, "step": 73020 }, { "grad_norm": 0.23896948993206024, "learning_rate": 2.0615577508639682e-06, "loss": 0.0095, "step": 73030 }, { "grad_norm": 0.24269908666610718, "learning_rate": 2.055688162025399e-06, "loss": 0.0117, "step": 73040 }, { "grad_norm": 0.44636332988739014, "learning_rate": 2.0498267655584546e-06, "loss": 0.0181, "step": 73050 }, { "grad_norm": 0.20680153369903564, "learning_rate": 2.0439735624646626e-06, "loss": 0.0101, "step": 73060 }, { "grad_norm": 0.3527849614620209, "learning_rate": 2.0381285537442008e-06, "loss": 0.0087, "step": 73070 }, { "grad_norm": 0.23026418685913086, "learning_rate": 2.032291740395803e-06, "loss": 0.0144, "step": 73080 }, { "grad_norm": 0.3866789937019348, "learning_rate": 2.0264631234168276e-06, "loss": 0.0157, "step": 73090 }, { "grad_norm": 0.267555832862854, "learning_rate": 2.0206427038032326e-06, "loss": 0.0141, "step": 73100 }, { "grad_norm": 0.3058982789516449, "learning_rate": 2.0148304825495457e-06, "loss": 0.0081, "step": 73110 }, { "grad_norm": 0.3317323923110962, "learning_rate": 2.0090264606489496e-06, "loss": 0.0075, "step": 73120 }, { "grad_norm": 0.365297794342041, "learning_rate": 2.003230639093162e-06, "loss": 0.0137, "step": 73130 }, { "grad_norm": 0.17740082740783691, "learning_rate": 1.9974430188725525e-06, "loss": 0.0165, "step": 73140 }, { "grad_norm": 0.396438330411911, "learning_rate": 1.9916636009760513e-06, "loss": 0.0129, "step": 73150 }, { "grad_norm": 0.25619661808013916, "learning_rate": 1.985892386391208e-06, "loss": 0.0142, "step": 73160 }, { "grad_norm": 0.15637293457984924, "learning_rate": 1.980129376104173e-06, "loss": 0.0128, "step": 73170 }, { "grad_norm": 0.13386906683444977, "learning_rate": 1.9743745710996796e-06, "loss": 0.0153, "step": 73180 }, { "grad_norm": 0.19728708267211914, "learning_rate": 1.9686279723610757e-06, "loss": 0.0114, "step": 73190 }, { "grad_norm": 0.4214472472667694, "learning_rate": 1.962889580870281e-06, "loss": 0.0117, "step": 73200 }, { "grad_norm": 0.19410328567028046, "learning_rate": 1.9571593976078385e-06, "loss": 0.0105, "step": 73210 }, { "grad_norm": 0.2810952961444855, "learning_rate": 1.9514374235528876e-06, "loss": 0.0125, "step": 73220 }, { "grad_norm": 0.5432409644126892, "learning_rate": 1.9457236596831408e-06, "loss": 0.0096, "step": 73230 }, { "grad_norm": 0.23407191038131714, "learning_rate": 1.940018106974939e-06, "loss": 0.0158, "step": 73240 }, { "grad_norm": 0.21072055399417877, "learning_rate": 1.9343207664032025e-06, "loss": 0.0094, "step": 73250 }, { "grad_norm": 0.12445542216300964, "learning_rate": 1.928631638941436e-06, "loss": 0.0071, "step": 73260 }, { "grad_norm": 0.2881241738796234, "learning_rate": 1.922950725561784e-06, "loss": 0.0094, "step": 73270 }, { "grad_norm": 0.20992498099803925, "learning_rate": 1.9172780272349312e-06, "loss": 0.0096, "step": 73280 }, { "grad_norm": 0.11637381464242935, "learning_rate": 1.911613544930202e-06, "loss": 0.014, "step": 73290 }, { "grad_norm": 0.21678729355335236, "learning_rate": 1.9059572796155e-06, "loss": 0.0128, "step": 73300 }, { "grad_norm": 0.3231964111328125, "learning_rate": 1.9003092322573192e-06, "loss": 0.0126, "step": 73310 }, { "grad_norm": 0.19420216977596283, "learning_rate": 1.8946694038207647e-06, "loss": 0.0127, "step": 73320 }, { "grad_norm": 0.6692609786987305, "learning_rate": 1.8890377952695215e-06, "loss": 0.0107, "step": 73330 }, { "grad_norm": 0.24195390939712524, "learning_rate": 1.8834144075658865e-06, "loss": 0.0116, "step": 73340 }, { "grad_norm": 0.26792821288108826, "learning_rate": 1.8777992416707357e-06, "loss": 0.012, "step": 73350 }, { "grad_norm": 0.22996127605438232, "learning_rate": 1.8721922985435458e-06, "loss": 0.0078, "step": 73360 }, { "grad_norm": 0.41175737977027893, "learning_rate": 1.8665935791424006e-06, "loss": 0.013, "step": 73370 }, { "grad_norm": 0.2666901648044586, "learning_rate": 1.8610030844239512e-06, "loss": 0.0104, "step": 73380 }, { "grad_norm": 0.561887800693512, "learning_rate": 1.8554208153434838e-06, "loss": 0.0134, "step": 73390 }, { "grad_norm": 0.14642742276191711, "learning_rate": 1.8498467728548296e-06, "loss": 0.0077, "step": 73400 }, { "grad_norm": 0.18920817971229553, "learning_rate": 1.8442809579104547e-06, "loss": 0.009, "step": 73410 }, { "grad_norm": 0.26747754216194153, "learning_rate": 1.8387233714614039e-06, "loss": 0.0135, "step": 73420 }, { "grad_norm": 0.204730823636055, "learning_rate": 1.8331740144573174e-06, "loss": 0.0137, "step": 73430 }, { "grad_norm": 0.24905148148536682, "learning_rate": 1.8276328878464255e-06, "loss": 0.011, "step": 73440 }, { "grad_norm": 0.21686534583568573, "learning_rate": 1.82209999257556e-06, "loss": 0.0112, "step": 73450 }, { "grad_norm": 0.1461462825536728, "learning_rate": 1.8165753295901311e-06, "loss": 0.0137, "step": 73460 }, { "grad_norm": 0.206600159406662, "learning_rate": 1.8110588998341616e-06, "loss": 0.0139, "step": 73470 }, { "grad_norm": 0.1788281798362732, "learning_rate": 1.805550704250253e-06, "loss": 0.0172, "step": 73480 }, { "grad_norm": 0.1339103728532791, "learning_rate": 1.8000507437796077e-06, "loss": 0.0073, "step": 73490 }, { "grad_norm": 0.1873745173215866, "learning_rate": 1.7945590193620242e-06, "loss": 0.0128, "step": 73500 }, { "grad_norm": 0.17506860196590424, "learning_rate": 1.7890755319358742e-06, "loss": 0.0137, "step": 73510 }, { "grad_norm": 0.38263943791389465, "learning_rate": 1.7836002824381525e-06, "loss": 0.015, "step": 73520 }, { "grad_norm": 0.22683869302272797, "learning_rate": 1.7781332718044165e-06, "loss": 0.0083, "step": 73530 }, { "grad_norm": 0.310580849647522, "learning_rate": 1.772674500968835e-06, "loss": 0.0142, "step": 73540 }, { "grad_norm": 0.23494598269462585, "learning_rate": 1.7672239708641624e-06, "loss": 0.0145, "step": 73550 }, { "grad_norm": 0.11777909100055695, "learning_rate": 1.7617816824217371e-06, "loss": 0.0106, "step": 73560 }, { "grad_norm": 0.3891746401786804, "learning_rate": 1.7563476365715148e-06, "loss": 0.0153, "step": 73570 }, { "grad_norm": 0.13282467424869537, "learning_rate": 1.750921834242003e-06, "loss": 0.0159, "step": 73580 }, { "grad_norm": 0.41171807050704956, "learning_rate": 1.7455042763603436e-06, "loss": 0.0129, "step": 73590 }, { "grad_norm": 0.16337160766124725, "learning_rate": 1.7400949638522345e-06, "loss": 0.0117, "step": 73600 }, { "grad_norm": 0.3276957869529724, "learning_rate": 1.734693897641987e-06, "loss": 0.0226, "step": 73610 }, { "grad_norm": 0.3813951015472412, "learning_rate": 1.7293010786524955e-06, "loss": 0.0113, "step": 73620 }, { "grad_norm": 0.3142051696777344, "learning_rate": 1.72391650780524e-06, "loss": 0.0111, "step": 73630 }, { "grad_norm": 0.27515608072280884, "learning_rate": 1.718540186020301e-06, "loss": 0.0137, "step": 73640 }, { "grad_norm": 0.4524767994880676, "learning_rate": 1.7131721142163437e-06, "loss": 0.0124, "step": 73650 }, { "grad_norm": 0.3056204617023468, "learning_rate": 1.7078122933106233e-06, "loss": 0.017, "step": 73660 }, { "grad_norm": 0.2264784723520279, "learning_rate": 1.7024607242189905e-06, "loss": 0.0129, "step": 73670 }, { "grad_norm": 0.2209123820066452, "learning_rate": 1.6971174078558749e-06, "loss": 0.0098, "step": 73680 }, { "grad_norm": 0.24242131412029266, "learning_rate": 1.6917823451343073e-06, "loss": 0.0133, "step": 73690 }, { "grad_norm": 0.46418073773384094, "learning_rate": 1.6864555369659142e-06, "loss": 0.0104, "step": 73700 }, { "grad_norm": 0.33585432171821594, "learning_rate": 1.681136984260878e-06, "loss": 0.0105, "step": 73710 }, { "grad_norm": 0.1382816731929779, "learning_rate": 1.6758266879280172e-06, "loss": 0.01, "step": 73720 }, { "grad_norm": 0.3158254027366638, "learning_rate": 1.6705246488746996e-06, "loss": 0.0105, "step": 73730 }, { "grad_norm": 0.12273597717285156, "learning_rate": 1.6652308680069062e-06, "loss": 0.0103, "step": 73740 }, { "grad_norm": 0.2522188425064087, "learning_rate": 1.6599453462292081e-06, "loss": 0.0112, "step": 73750 }, { "grad_norm": 0.5073825120925903, "learning_rate": 1.6546680844447326e-06, "loss": 0.0175, "step": 73760 }, { "grad_norm": 0.2549061179161072, "learning_rate": 1.6493990835552475e-06, "loss": 0.0131, "step": 73770 }, { "grad_norm": 0.13383044302463531, "learning_rate": 1.6441383444610547e-06, "loss": 0.0098, "step": 73780 }, { "grad_norm": 0.1848829984664917, "learning_rate": 1.6388858680610908e-06, "loss": 0.0164, "step": 73790 }, { "grad_norm": 0.29309579730033875, "learning_rate": 1.6336416552528543e-06, "loss": 0.0129, "step": 73800 }, { "grad_norm": 0.27437734603881836, "learning_rate": 1.6284057069324288e-06, "loss": 0.0141, "step": 73810 }, { "grad_norm": 0.16162872314453125, "learning_rate": 1.623178023994504e-06, "loss": 0.0188, "step": 73820 }, { "grad_norm": 0.40263059735298157, "learning_rate": 1.6179586073323483e-06, "loss": 0.015, "step": 73830 }, { "grad_norm": 0.28690028190612793, "learning_rate": 1.612747457837821e-06, "loss": 0.0101, "step": 73840 }, { "grad_norm": 0.16738547384738922, "learning_rate": 1.607544576401354e-06, "loss": 0.0088, "step": 73850 }, { "grad_norm": 0.271152526140213, "learning_rate": 1.6023499639119754e-06, "loss": 0.0136, "step": 73860 }, { "grad_norm": 0.27357929944992065, "learning_rate": 1.5971636212573138e-06, "loss": 0.0104, "step": 73870 }, { "grad_norm": 0.2620633542537689, "learning_rate": 1.5919855493235714e-06, "loss": 0.0103, "step": 73880 }, { "grad_norm": 0.1459319293498993, "learning_rate": 1.586815748995535e-06, "loss": 0.0091, "step": 73890 }, { "grad_norm": 0.4065585434436798, "learning_rate": 1.5816542211565866e-06, "loss": 0.0148, "step": 73900 }, { "grad_norm": 0.2496352642774582, "learning_rate": 1.5765009666886765e-06, "loss": 0.0129, "step": 73910 }, { "grad_norm": 0.276203453540802, "learning_rate": 1.5713559864723781e-06, "loss": 0.0145, "step": 73920 }, { "grad_norm": 0.2052425742149353, "learning_rate": 1.5662192813868104e-06, "loss": 0.0123, "step": 73930 }, { "grad_norm": 0.29709795117378235, "learning_rate": 1.561090852309699e-06, "loss": 0.0171, "step": 73940 }, { "grad_norm": 0.23188097774982452, "learning_rate": 1.5559707001173651e-06, "loss": 0.009, "step": 73950 }, { "grad_norm": 0.3100926876068115, "learning_rate": 1.5508588256846757e-06, "loss": 0.0131, "step": 73960 }, { "grad_norm": 0.22752143442630768, "learning_rate": 1.5457552298851319e-06, "loss": 0.0136, "step": 73970 }, { "grad_norm": 0.22685159742832184, "learning_rate": 1.5406599135907918e-06, "loss": 0.0089, "step": 73980 }, { "grad_norm": 0.4284529685974121, "learning_rate": 1.5355728776723088e-06, "loss": 0.0104, "step": 73990 }, { "grad_norm": 0.15568508207798004, "learning_rate": 1.5304941229989155e-06, "loss": 0.0117, "step": 74000 }, { "grad_norm": 0.38854488730430603, "learning_rate": 1.5254236504384345e-06, "loss": 0.0123, "step": 74010 }, { "grad_norm": 0.19219321012496948, "learning_rate": 1.5203614608572726e-06, "loss": 0.0139, "step": 74020 }, { "grad_norm": 0.2080501914024353, "learning_rate": 1.5153075551204044e-06, "loss": 0.0121, "step": 74030 }, { "grad_norm": 0.1950252652168274, "learning_rate": 1.5102619340914225e-06, "loss": 0.0215, "step": 74040 }, { "grad_norm": 0.4119332432746887, "learning_rate": 1.505224598632482e-06, "loss": 0.0106, "step": 74050 }, { "grad_norm": 0.2163918912410736, "learning_rate": 1.5001955496043164e-06, "loss": 0.017, "step": 74060 }, { "grad_norm": 0.3823372423648834, "learning_rate": 1.4951747878662604e-06, "loss": 0.0145, "step": 74070 }, { "grad_norm": 0.13823513686656952, "learning_rate": 1.4901623142762221e-06, "loss": 0.0112, "step": 74080 }, { "grad_norm": 0.21516825258731842, "learning_rate": 1.4851581296907001e-06, "loss": 0.0156, "step": 74090 }, { "grad_norm": 0.17678464949131012, "learning_rate": 1.4801622349647714e-06, "loss": 0.0071, "step": 74100 }, { "grad_norm": 0.18551866710186005, "learning_rate": 1.4751746309520976e-06, "loss": 0.0139, "step": 74110 }, { "grad_norm": 0.1536477655172348, "learning_rate": 1.4701953185049132e-06, "loss": 0.0155, "step": 74120 }, { "grad_norm": 0.15530292689800262, "learning_rate": 1.4652242984740661e-06, "loss": 0.0177, "step": 74130 }, { "grad_norm": 0.20509101450443268, "learning_rate": 1.4602615717089484e-06, "loss": 0.017, "step": 74140 }, { "grad_norm": 0.2782590389251709, "learning_rate": 1.4553071390575657e-06, "loss": 0.0153, "step": 74150 }, { "grad_norm": 0.19424310326576233, "learning_rate": 1.450361001366496e-06, "loss": 0.0162, "step": 74160 }, { "grad_norm": 0.2660462260246277, "learning_rate": 1.4454231594809021e-06, "loss": 0.014, "step": 74170 }, { "grad_norm": 0.3241432011127472, "learning_rate": 1.4404936142445036e-06, "loss": 0.0107, "step": 74180 }, { "grad_norm": 0.29189762473106384, "learning_rate": 1.4355723664996546e-06, "loss": 0.0104, "step": 74190 }, { "grad_norm": 0.33067798614501953, "learning_rate": 1.4306594170872433e-06, "loss": 0.0125, "step": 74200 }, { "grad_norm": 0.29423049092292786, "learning_rate": 1.4257547668467598e-06, "loss": 0.0164, "step": 74210 }, { "grad_norm": 0.38965991139411926, "learning_rate": 1.4208584166162886e-06, "loss": 0.0111, "step": 74220 }, { "grad_norm": 0.27708980441093445, "learning_rate": 1.4159703672324553e-06, "loss": 0.0111, "step": 74230 }, { "grad_norm": 0.15981455147266388, "learning_rate": 1.4110906195305252e-06, "loss": 0.0124, "step": 74240 }, { "grad_norm": 0.2779882848262787, "learning_rate": 1.406219174344292e-06, "loss": 0.0142, "step": 74250 }, { "grad_norm": 0.19436219334602356, "learning_rate": 1.401356032506157e-06, "loss": 0.0122, "step": 74260 }, { "grad_norm": 0.10851608216762543, "learning_rate": 1.3965011948471051e-06, "loss": 0.0186, "step": 74270 }, { "grad_norm": 0.40112361311912537, "learning_rate": 1.3916546621966842e-06, "loss": 0.0137, "step": 74280 }, { "grad_norm": 0.22395049035549164, "learning_rate": 1.3868164353830482e-06, "loss": 0.0108, "step": 74290 }, { "grad_norm": 0.39513638615608215, "learning_rate": 1.3819865152329082e-06, "loss": 0.0102, "step": 74300 }, { "grad_norm": 0.14107921719551086, "learning_rate": 1.377164902571565e-06, "loss": 0.0179, "step": 74310 }, { "grad_norm": 0.21853916347026825, "learning_rate": 1.372351598222904e-06, "loss": 0.0128, "step": 74320 }, { "grad_norm": 0.22278833389282227, "learning_rate": 1.3675466030093843e-06, "loss": 0.0106, "step": 74330 }, { "grad_norm": 0.16621996462345123, "learning_rate": 1.3627499177520486e-06, "loss": 0.0162, "step": 74340 }, { "grad_norm": 0.16598325967788696, "learning_rate": 1.3579615432705196e-06, "loss": 0.0109, "step": 74350 }, { "grad_norm": 0.4333379566669464, "learning_rate": 1.3531814803829978e-06, "loss": 0.0155, "step": 74360 }, { "grad_norm": 0.1804477870464325, "learning_rate": 1.3484097299062803e-06, "loss": 0.009, "step": 74370 }, { "grad_norm": 0.1317485123872757, "learning_rate": 1.343646292655698e-06, "loss": 0.0115, "step": 74380 }, { "grad_norm": 0.2708349823951721, "learning_rate": 1.3388911694452277e-06, "loss": 0.0143, "step": 74390 }, { "grad_norm": 0.46333634853363037, "learning_rate": 1.3341443610873638e-06, "loss": 0.0173, "step": 74400 }, { "grad_norm": 0.28740033507347107, "learning_rate": 1.329405868393213e-06, "loss": 0.0106, "step": 74410 }, { "grad_norm": 0.25335779786109924, "learning_rate": 1.3246756921724613e-06, "loss": 0.0166, "step": 74420 }, { "grad_norm": 0.21124167740345, "learning_rate": 1.3199538332333506e-06, "loss": 0.0123, "step": 74430 }, { "grad_norm": 0.6587502360343933, "learning_rate": 1.3152402923827411e-06, "loss": 0.0159, "step": 74440 }, { "grad_norm": 0.24107052385807037, "learning_rate": 1.3105350704260277e-06, "loss": 0.0115, "step": 74450 }, { "grad_norm": 0.31740570068359375, "learning_rate": 1.305838168167206e-06, "loss": 0.0101, "step": 74460 }, { "grad_norm": 0.26397812366485596, "learning_rate": 1.301149586408862e-06, "loss": 0.0089, "step": 74470 }, { "grad_norm": 0.15815460681915283, "learning_rate": 1.2964693259521322e-06, "loss": 0.0144, "step": 74480 }, { "grad_norm": 0.5032587647438049, "learning_rate": 1.2917973875967548e-06, "loss": 0.0134, "step": 74490 }, { "grad_norm": 0.3507850766181946, "learning_rate": 1.2871337721410249e-06, "loss": 0.0128, "step": 74500 }, { "grad_norm": 0.199195995926857, "learning_rate": 1.2824784803818379e-06, "loss": 0.0135, "step": 74510 }, { "grad_norm": 0.30468326807022095, "learning_rate": 1.2778315131146467e-06, "loss": 0.009, "step": 74520 }, { "grad_norm": 0.3176724314689636, "learning_rate": 1.2731928711334994e-06, "loss": 0.0112, "step": 74530 }, { "grad_norm": 0.2124495655298233, "learning_rate": 1.2685625552310122e-06, "loss": 0.01, "step": 74540 }, { "grad_norm": 0.16395746171474457, "learning_rate": 1.263940566198374e-06, "loss": 0.0109, "step": 74550 }, { "grad_norm": 0.18922941386699677, "learning_rate": 1.259326904825353e-06, "loss": 0.0121, "step": 74560 }, { "grad_norm": 0.36248448491096497, "learning_rate": 1.2547215719003137e-06, "loss": 0.0117, "step": 74570 }, { "grad_norm": 0.24223355948925018, "learning_rate": 1.2501245682101703e-06, "loss": 0.0113, "step": 74580 }, { "grad_norm": 0.2253388911485672, "learning_rate": 1.2455358945404171e-06, "loss": 0.0097, "step": 74590 }, { "grad_norm": 0.0665455088019371, "learning_rate": 1.2409555516751493e-06, "loss": 0.0078, "step": 74600 }, { "grad_norm": 0.23076790571212769, "learning_rate": 1.2363835403970125e-06, "loss": 0.0119, "step": 74610 }, { "grad_norm": 0.26802217960357666, "learning_rate": 1.2318198614872489e-06, "loss": 0.0141, "step": 74620 }, { "grad_norm": 0.3171154856681824, "learning_rate": 1.2272645157256457e-06, "loss": 0.011, "step": 74630 }, { "grad_norm": 0.2644890546798706, "learning_rate": 1.222717503890608e-06, "loss": 0.0187, "step": 74640 }, { "grad_norm": 0.5198145508766174, "learning_rate": 1.218178826759081e-06, "loss": 0.0142, "step": 74650 }, { "grad_norm": 0.4275026321411133, "learning_rate": 1.213648485106611e-06, "loss": 0.0139, "step": 74660 }, { "grad_norm": 0.1774950474500656, "learning_rate": 1.2091264797073066e-06, "loss": 0.0129, "step": 74670 }, { "grad_norm": 0.3098706305027008, "learning_rate": 1.2046128113338494e-06, "loss": 0.0249, "step": 74680 }, { "grad_norm": 0.9016554951667786, "learning_rate": 1.200107480757512e-06, "loss": 0.0088, "step": 74690 }, { "grad_norm": 0.1616395264863968, "learning_rate": 1.1956104887481168e-06, "loss": 0.02, "step": 74700 }, { "grad_norm": 0.39729952812194824, "learning_rate": 1.191121836074094e-06, "loss": 0.0104, "step": 74710 }, { "grad_norm": 0.17150862514972687, "learning_rate": 1.1866415235024186e-06, "loss": 0.0098, "step": 74720 }, { "grad_norm": 0.2359112948179245, "learning_rate": 1.182169551798662e-06, "loss": 0.0151, "step": 74730 }, { "grad_norm": 0.34498360753059387, "learning_rate": 1.177705921726957e-06, "loss": 0.0116, "step": 74740 }, { "grad_norm": 0.076769158244133, "learning_rate": 1.173250634050016e-06, "loss": 0.0105, "step": 74750 }, { "grad_norm": 0.2118915170431137, "learning_rate": 1.1688036895291354e-06, "loss": 0.012, "step": 74760 }, { "grad_norm": 0.2976742684841156, "learning_rate": 1.1643650889241574e-06, "loss": 0.0155, "step": 74770 }, { "grad_norm": 0.24155493080615997, "learning_rate": 1.1599348329935311e-06, "loss": 0.0108, "step": 74780 }, { "grad_norm": 0.1543681025505066, "learning_rate": 1.1555129224942673e-06, "loss": 0.0189, "step": 74790 }, { "grad_norm": 0.17951291799545288, "learning_rate": 1.1510993581819396e-06, "loss": 0.0167, "step": 74800 }, { "grad_norm": 0.17710916697978973, "learning_rate": 1.1466941408107112e-06, "loss": 0.0212, "step": 74810 }, { "grad_norm": 0.23467202484607697, "learning_rate": 1.1422972711333247e-06, "loss": 0.0107, "step": 74820 }, { "grad_norm": 0.2206815928220749, "learning_rate": 1.1379087499010565e-06, "loss": 0.0092, "step": 74830 }, { "grad_norm": 0.29004427790641785, "learning_rate": 1.1335285778638127e-06, "loss": 0.01, "step": 74840 }, { "grad_norm": 0.20727625489234924, "learning_rate": 1.1291567557700333e-06, "loss": 0.0157, "step": 74850 }, { "grad_norm": 0.23809634149074554, "learning_rate": 1.124793284366743e-06, "loss": 0.0145, "step": 74860 }, { "grad_norm": 0.30614885687828064, "learning_rate": 1.1204381643995455e-06, "loss": 0.0117, "step": 74870 }, { "grad_norm": 0.24632009863853455, "learning_rate": 1.116091396612595e-06, "loss": 0.0137, "step": 74880 }, { "grad_norm": 0.17974361777305603, "learning_rate": 1.1117529817486538e-06, "loss": 0.0132, "step": 74890 }, { "grad_norm": 0.7805415391921997, "learning_rate": 1.107422920549034e-06, "loss": 0.0104, "step": 74900 }, { "grad_norm": 0.28745195269584656, "learning_rate": 1.1031012137536157e-06, "loss": 0.0135, "step": 74910 }, { "grad_norm": 0.3330689072608948, "learning_rate": 1.0987878621008695e-06, "loss": 0.0129, "step": 74920 }, { "grad_norm": 0.3557327091693878, "learning_rate": 1.0944828663278284e-06, "loss": 0.011, "step": 74930 }, { "grad_norm": 0.12021856755018234, "learning_rate": 1.0901862271700925e-06, "loss": 0.0134, "step": 74940 }, { "grad_norm": 0.18096667528152466, "learning_rate": 1.0858979453618467e-06, "loss": 0.0117, "step": 74950 }, { "grad_norm": 0.40171369910240173, "learning_rate": 1.0816180216358441e-06, "loss": 0.013, "step": 74960 }, { "grad_norm": 0.3065490126609802, "learning_rate": 1.0773464567233937e-06, "loss": 0.0123, "step": 74970 }, { "grad_norm": 0.27143874764442444, "learning_rate": 1.0730832513543953e-06, "loss": 0.0123, "step": 74980 }, { "grad_norm": 0.27878427505493164, "learning_rate": 1.0688284062573218e-06, "loss": 0.0163, "step": 74990 }, { "grad_norm": 0.20728106796741486, "learning_rate": 1.0645819221591969e-06, "loss": 0.0078, "step": 75000 }, { "grad_norm": 0.34305539727211, "learning_rate": 1.0603437997856347e-06, "loss": 0.0134, "step": 75010 }, { "grad_norm": 0.32150107622146606, "learning_rate": 1.0561140398608228e-06, "loss": 0.0124, "step": 75020 }, { "grad_norm": 0.4507812261581421, "learning_rate": 1.0518926431074937e-06, "loss": 0.0101, "step": 75030 }, { "grad_norm": 0.16944174468517303, "learning_rate": 1.0476796102469877e-06, "loss": 0.0126, "step": 75040 }, { "grad_norm": 0.181356281042099, "learning_rate": 1.0434749419991786e-06, "loss": 0.0115, "step": 75050 }, { "grad_norm": 0.2985464334487915, "learning_rate": 1.0392786390825415e-06, "loss": 0.011, "step": 75060 }, { "grad_norm": 0.30054426193237305, "learning_rate": 1.0350907022141087e-06, "loss": 0.0164, "step": 75070 }, { "grad_norm": 0.6273012757301331, "learning_rate": 1.0309111321094744e-06, "loss": 0.0134, "step": 75080 }, { "grad_norm": 0.22922804951667786, "learning_rate": 1.026739929482834e-06, "loss": 0.0089, "step": 75090 }, { "grad_norm": 0.4273987114429474, "learning_rate": 1.022577095046906e-06, "loss": 0.0131, "step": 75100 }, { "grad_norm": 0.2085106521844864, "learning_rate": 1.0184226295130217e-06, "loss": 0.0097, "step": 75110 }, { "grad_norm": 0.2990005314350128, "learning_rate": 1.0142765335910575e-06, "loss": 0.0131, "step": 75120 }, { "grad_norm": 0.27825990319252014, "learning_rate": 1.0101388079894746e-06, "loss": 0.0117, "step": 75130 }, { "grad_norm": 0.20410126447677612, "learning_rate": 1.0060094534152908e-06, "loss": 0.0118, "step": 75140 }, { "grad_norm": 0.23259307444095612, "learning_rate": 1.0018884705741028e-06, "loss": 0.0108, "step": 75150 }, { "grad_norm": 0.18846389651298523, "learning_rate": 9.977758601700805e-07, "loss": 0.0172, "step": 75160 }, { "grad_norm": 0.1919163167476654, "learning_rate": 9.93671622905945e-07, "loss": 0.0198, "step": 75170 }, { "grad_norm": 0.25535905361175537, "learning_rate": 9.895757594830024e-07, "loss": 0.0118, "step": 75180 }, { "grad_norm": 0.24110472202301025, "learning_rate": 9.854882706011204e-07, "loss": 0.0104, "step": 75190 }, { "grad_norm": 0.40923669934272766, "learning_rate": 9.814091569587513e-07, "loss": 0.0109, "step": 75200 }, { "grad_norm": 0.30321601033210754, "learning_rate": 9.77338419252888e-07, "loss": 0.0191, "step": 75210 }, { "grad_norm": 0.3811969459056854, "learning_rate": 9.732760581791289e-07, "loss": 0.0143, "step": 75220 }, { "grad_norm": 0.24063512682914734, "learning_rate": 9.692220744315972e-07, "loss": 0.0112, "step": 75230 }, { "grad_norm": 0.22339513897895813, "learning_rate": 9.651764687030162e-07, "loss": 0.0108, "step": 75240 }, { "grad_norm": 0.15726101398468018, "learning_rate": 9.611392416846776e-07, "loss": 0.0093, "step": 75250 }, { "grad_norm": 0.27167272567749023, "learning_rate": 9.571103940664238e-07, "loss": 0.0135, "step": 75260 }, { "grad_norm": 0.20177409052848816, "learning_rate": 9.530899265366822e-07, "loss": 0.0133, "step": 75270 }, { "grad_norm": 0.17931067943572998, "learning_rate": 9.490778397824307e-07, "loss": 0.0108, "step": 75280 }, { "grad_norm": 0.33696019649505615, "learning_rate": 9.450741344892322e-07, "loss": 0.0108, "step": 75290 }, { "grad_norm": 0.30881214141845703, "learning_rate": 9.410788113412116e-07, "loss": 0.0106, "step": 75300 }, { "grad_norm": 0.6116241216659546, "learning_rate": 9.370918710210563e-07, "loss": 0.0111, "step": 75310 }, { "grad_norm": 0.22103355824947357, "learning_rate": 9.331133142100268e-07, "loss": 0.0123, "step": 75320 }, { "grad_norm": 0.21326178312301636, "learning_rate": 9.291431415879459e-07, "loss": 0.0088, "step": 75330 }, { "grad_norm": 0.4394797086715698, "learning_rate": 9.251813538332155e-07, "loss": 0.015, "step": 75340 }, { "grad_norm": 0.13849817216396332, "learning_rate": 9.212279516227884e-07, "loss": 0.0111, "step": 75350 }, { "grad_norm": 0.35452407598495483, "learning_rate": 9.172829356322021e-07, "loss": 0.0103, "step": 75360 }, { "grad_norm": 0.38704243302345276, "learning_rate": 9.133463065355452e-07, "loss": 0.0121, "step": 75370 }, { "grad_norm": 0.2110288143157959, "learning_rate": 9.094180650054796e-07, "loss": 0.0129, "step": 75380 }, { "grad_norm": 0.3223115801811218, "learning_rate": 9.054982117132404e-07, "loss": 0.0093, "step": 75390 }, { "grad_norm": 0.2818225920200348, "learning_rate": 9.015867473286143e-07, "loss": 0.0128, "step": 75400 }, { "grad_norm": 0.488201767206192, "learning_rate": 8.976836725199778e-07, "loss": 0.0124, "step": 75410 }, { "grad_norm": 0.20810094475746155, "learning_rate": 8.937889879542416e-07, "loss": 0.0106, "step": 75420 }, { "grad_norm": 0.5724019408226013, "learning_rate": 8.899026942969068e-07, "loss": 0.0116, "step": 75430 }, { "grad_norm": 0.2774491310119629, "learning_rate": 8.860247922120424e-07, "loss": 0.0132, "step": 75440 }, { "grad_norm": 0.2862516939640045, "learning_rate": 8.821552823622737e-07, "loss": 0.0109, "step": 75450 }, { "grad_norm": 0.16375315189361572, "learning_rate": 8.782941654087884e-07, "loss": 0.0107, "step": 75460 }, { "grad_norm": 0.6146498918533325, "learning_rate": 8.744414420113478e-07, "loss": 0.0146, "step": 75470 }, { "grad_norm": 0.15413960814476013, "learning_rate": 8.705971128282753e-07, "loss": 0.0099, "step": 75480 }, { "grad_norm": 0.10319793969392776, "learning_rate": 8.66761178516473e-07, "loss": 0.0078, "step": 75490 }, { "grad_norm": 0.21315954625606537, "learning_rate": 8.629336397313781e-07, "loss": 0.0096, "step": 75500 }, { "grad_norm": 0.23893696069717407, "learning_rate": 8.591144971270282e-07, "loss": 0.0113, "step": 75510 }, { "grad_norm": 0.15283407270908356, "learning_rate": 8.553037513560069e-07, "loss": 0.0091, "step": 75520 }, { "grad_norm": 0.3582471013069153, "learning_rate": 8.515014030694546e-07, "loss": 0.0107, "step": 75530 }, { "grad_norm": 0.15275399386882782, "learning_rate": 8.47707452917107e-07, "loss": 0.0152, "step": 75540 }, { "grad_norm": 0.2979365587234497, "learning_rate": 8.439219015472343e-07, "loss": 0.0134, "step": 75550 }, { "grad_norm": 0.4248731732368469, "learning_rate": 8.401447496066861e-07, "loss": 0.0147, "step": 75560 }, { "grad_norm": 0.2706786096096039, "learning_rate": 8.363759977408792e-07, "loss": 0.01, "step": 75570 }, { "grad_norm": 0.15361735224723816, "learning_rate": 8.326156465937817e-07, "loss": 0.0125, "step": 75580 }, { "grad_norm": 0.48640307784080505, "learning_rate": 8.28863696807941e-07, "loss": 0.0171, "step": 75590 }, { "grad_norm": 0.20969286561012268, "learning_rate": 8.251201490244609e-07, "loss": 0.0112, "step": 75600 }, { "grad_norm": 0.308030903339386, "learning_rate": 8.213850038830129e-07, "loss": 0.0096, "step": 75610 }, { "grad_norm": 0.13639356195926666, "learning_rate": 8.176582620218309e-07, "loss": 0.0122, "step": 75620 }, { "grad_norm": 0.2917368710041046, "learning_rate": 8.139399240777057e-07, "loss": 0.0131, "step": 75630 }, { "grad_norm": 0.30511674284935, "learning_rate": 8.102299906860122e-07, "loss": 0.0157, "step": 75640 }, { "grad_norm": 0.4036498963832855, "learning_rate": 8.065284624806657e-07, "loss": 0.0119, "step": 75650 }, { "grad_norm": 0.4373182952404022, "learning_rate": 8.028353400941601e-07, "loss": 0.013, "step": 75660 }, { "grad_norm": 0.20528054237365723, "learning_rate": 7.991506241575575e-07, "loss": 0.0124, "step": 75670 }, { "grad_norm": 0.34557247161865234, "learning_rate": 7.954743153004541e-07, "loss": 0.0147, "step": 75680 }, { "grad_norm": 0.23155342042446136, "learning_rate": 7.918064141510528e-07, "loss": 0.012, "step": 75690 }, { "grad_norm": 0.2776011824607849, "learning_rate": 7.881469213360859e-07, "loss": 0.0126, "step": 75700 }, { "grad_norm": 0.3239346146583557, "learning_rate": 7.844958374808642e-07, "loss": 0.0125, "step": 75710 }, { "grad_norm": 0.21924418210983276, "learning_rate": 7.808531632092608e-07, "loss": 0.0095, "step": 75720 }, { "grad_norm": 0.2072480320930481, "learning_rate": 7.772188991436946e-07, "loss": 0.0105, "step": 75730 }, { "grad_norm": 0.237526997923851, "learning_rate": 7.735930459051799e-07, "loss": 0.0112, "step": 75740 }, { "grad_norm": 0.23949851095676422, "learning_rate": 7.699756041132655e-07, "loss": 0.0125, "step": 75750 }, { "grad_norm": 0.21919891238212585, "learning_rate": 7.663665743860793e-07, "loss": 0.0105, "step": 75760 }, { "grad_norm": 0.32515060901641846, "learning_rate": 7.627659573403001e-07, "loss": 0.0135, "step": 75770 }, { "grad_norm": 0.218545600771904, "learning_rate": 7.591737535911802e-07, "loss": 0.0098, "step": 75780 }, { "grad_norm": 0.4529820680618286, "learning_rate": 7.555899637525288e-07, "loss": 0.0126, "step": 75790 }, { "grad_norm": 0.30212777853012085, "learning_rate": 7.520145884367058e-07, "loss": 0.0108, "step": 75800 }, { "grad_norm": 0.07285809516906738, "learning_rate": 7.484476282546615e-07, "loss": 0.0115, "step": 75810 }, { "grad_norm": 0.21955883502960205, "learning_rate": 7.448890838158806e-07, "loss": 0.0167, "step": 75820 }, { "grad_norm": 0.32316750288009644, "learning_rate": 7.413389557284267e-07, "loss": 0.0123, "step": 75830 }, { "grad_norm": 0.32596129179000854, "learning_rate": 7.377972445989201e-07, "loss": 0.0114, "step": 75840 }, { "grad_norm": 0.15574660897254944, "learning_rate": 7.342639510325377e-07, "loss": 0.0124, "step": 75850 }, { "grad_norm": 0.3257290720939636, "learning_rate": 7.307390756330246e-07, "loss": 0.0148, "step": 75860 }, { "grad_norm": 0.36893799901008606, "learning_rate": 7.272226190026876e-07, "loss": 0.01, "step": 75870 }, { "grad_norm": 0.2769981026649475, "learning_rate": 7.237145817423907e-07, "loss": 0.0106, "step": 75880 }, { "grad_norm": 0.7232184410095215, "learning_rate": 7.202149644515654e-07, "loss": 0.0125, "step": 75890 }, { "grad_norm": 0.347434937953949, "learning_rate": 7.167237677281946e-07, "loss": 0.0128, "step": 75900 }, { "grad_norm": 0.24575674533843994, "learning_rate": 7.132409921688288e-07, "loss": 0.0094, "step": 75910 }, { "grad_norm": 0.23457013070583344, "learning_rate": 7.09766638368592e-07, "loss": 0.0078, "step": 75920 }, { "grad_norm": 0.10369335860013962, "learning_rate": 7.063007069211313e-07, "loss": 0.013, "step": 75930 }, { "grad_norm": 0.3022695481777191, "learning_rate": 7.028431984187067e-07, "loss": 0.0097, "step": 75940 }, { "grad_norm": 0.22504375874996185, "learning_rate": 6.993941134520898e-07, "loss": 0.0134, "step": 75950 }, { "grad_norm": 0.1227872371673584, "learning_rate": 6.959534526106537e-07, "loss": 0.0109, "step": 75960 }, { "grad_norm": 0.2470698058605194, "learning_rate": 6.925212164822948e-07, "loss": 0.0102, "step": 75970 }, { "grad_norm": 0.24783970415592194, "learning_rate": 6.890974056535049e-07, "loss": 0.0147, "step": 75980 }, { "grad_norm": 0.1971515566110611, "learning_rate": 6.856820207093106e-07, "loss": 0.0116, "step": 75990 }, { "grad_norm": 0.2219456136226654, "learning_rate": 6.822750622333063e-07, "loss": 0.0099, "step": 76000 }, { "grad_norm": 1.1454474925994873, "learning_rate": 6.788765308076539e-07, "loss": 0.0167, "step": 76010 }, { "grad_norm": 0.2335021197795868, "learning_rate": 6.754864270130668e-07, "loss": 0.0117, "step": 76020 }, { "grad_norm": 0.5466521382331848, "learning_rate": 6.721047514288203e-07, "loss": 0.0113, "step": 76030 }, { "grad_norm": 0.2505655288696289, "learning_rate": 6.687315046327469e-07, "loss": 0.0129, "step": 76040 }, { "grad_norm": 0.2663341462612152, "learning_rate": 6.65366687201252e-07, "loss": 0.0129, "step": 76050 }, { "grad_norm": 0.4444893002510071, "learning_rate": 6.620102997092814e-07, "loss": 0.0126, "step": 76060 }, { "grad_norm": 0.32808029651641846, "learning_rate": 6.586623427303596e-07, "loss": 0.0089, "step": 76070 }, { "grad_norm": 0.09571105986833572, "learning_rate": 6.553228168365455e-07, "loss": 0.013, "step": 76080 }, { "grad_norm": 0.3352401852607727, "learning_rate": 6.519917225984884e-07, "loss": 0.0182, "step": 76090 }, { "grad_norm": 0.3634765148162842, "learning_rate": 6.486690605853718e-07, "loss": 0.0135, "step": 76100 }, { "grad_norm": 0.36412227153778076, "learning_rate": 6.453548313649527e-07, "loss": 0.0102, "step": 76110 }, { "grad_norm": 0.39217692613601685, "learning_rate": 6.420490355035446e-07, "loss": 0.0112, "step": 76120 }, { "grad_norm": 0.17513716220855713, "learning_rate": 6.387516735660071e-07, "loss": 0.012, "step": 76130 }, { "grad_norm": 0.5136923789978027, "learning_rate": 6.35462746115778e-07, "loss": 0.0106, "step": 76140 }, { "grad_norm": 0.10895051807165146, "learning_rate": 6.321822537148358e-07, "loss": 0.0115, "step": 76150 }, { "grad_norm": 0.14730966091156006, "learning_rate": 6.289101969237432e-07, "loss": 0.0136, "step": 76160 }, { "grad_norm": 0.41450539231300354, "learning_rate": 6.256465763015918e-07, "loss": 0.0116, "step": 76170 }, { "grad_norm": 0.19761580228805542, "learning_rate": 6.22391392406052e-07, "loss": 0.0113, "step": 76180 }, { "grad_norm": 0.15342137217521667, "learning_rate": 6.191446457933403e-07, "loss": 0.0079, "step": 76190 }, { "grad_norm": 0.7135657072067261, "learning_rate": 6.159063370182406e-07, "loss": 0.0109, "step": 76200 }, { "grad_norm": 0.3315909802913666, "learning_rate": 6.126764666340879e-07, "loss": 0.0098, "step": 76210 }, { "grad_norm": 0.29377469420433044, "learning_rate": 6.094550351927852e-07, "loss": 0.0122, "step": 76220 }, { "grad_norm": 0.2734772861003876, "learning_rate": 6.062420432447757e-07, "loss": 0.0131, "step": 76230 }, { "grad_norm": 0.17554089426994324, "learning_rate": 6.030374913390813e-07, "loss": 0.0116, "step": 76240 }, { "grad_norm": 0.36098599433898926, "learning_rate": 5.998413800232694e-07, "loss": 0.0125, "step": 76250 }, { "grad_norm": 0.22160513699054718, "learning_rate": 5.966537098434755e-07, "loss": 0.0104, "step": 76260 }, { "grad_norm": 0.2977132201194763, "learning_rate": 5.934744813443694e-07, "loss": 0.0111, "step": 76270 }, { "grad_norm": 0.3114241063594818, "learning_rate": 5.903036950692054e-07, "loss": 0.0113, "step": 76280 }, { "grad_norm": 0.35135528445243835, "learning_rate": 5.871413515597835e-07, "loss": 0.0163, "step": 76290 }, { "grad_norm": 0.24463117122650146, "learning_rate": 5.839874513564547e-07, "loss": 0.0087, "step": 76300 }, { "grad_norm": 0.21159140765666962, "learning_rate": 5.808419949981436e-07, "loss": 0.0139, "step": 76310 }, { "grad_norm": 0.21298719942569733, "learning_rate": 5.777049830223257e-07, "loss": 0.0107, "step": 76320 }, { "grad_norm": 0.2668173015117645, "learning_rate": 5.745764159650114e-07, "loss": 0.0099, "step": 76330 }, { "grad_norm": 0.2236642837524414, "learning_rate": 5.71456294360806e-07, "loss": 0.0112, "step": 76340 }, { "grad_norm": 0.35818955302238464, "learning_rate": 5.683446187428443e-07, "loss": 0.015, "step": 76350 }, { "grad_norm": 0.2227257788181305, "learning_rate": 5.652413896428288e-07, "loss": 0.0101, "step": 76360 }, { "grad_norm": 0.264687180519104, "learning_rate": 5.621466075910131e-07, "loss": 0.0108, "step": 76370 }, { "grad_norm": 0.1363338679075241, "learning_rate": 5.590602731162187e-07, "loss": 0.0118, "step": 76380 }, { "grad_norm": 0.4027367830276489, "learning_rate": 5.559823867458125e-07, "loss": 0.0098, "step": 76390 }, { "grad_norm": 0.15416209399700165, "learning_rate": 5.529129490057128e-07, "loss": 0.0167, "step": 76400 }, { "grad_norm": 0.18093852698802948, "learning_rate": 5.498519604204167e-07, "loss": 0.012, "step": 76410 }, { "grad_norm": 0.3161696493625641, "learning_rate": 5.467994215129501e-07, "loss": 0.0097, "step": 76420 }, { "grad_norm": 0.3636932373046875, "learning_rate": 5.437553328049183e-07, "loss": 0.013, "step": 76430 }, { "grad_norm": 0.16859199106693268, "learning_rate": 5.407196948164661e-07, "loss": 0.0104, "step": 76440 }, { "grad_norm": 0.22619423270225525, "learning_rate": 5.376925080663009e-07, "loss": 0.0134, "step": 76450 }, { "grad_norm": 0.21141426265239716, "learning_rate": 5.346737730716977e-07, "loss": 0.0127, "step": 76460 }, { "grad_norm": 0.19092874228954315, "learning_rate": 5.316634903484607e-07, "loss": 0.0086, "step": 76470 }, { "grad_norm": 0.4439754784107208, "learning_rate": 5.286616604109728e-07, "loss": 0.0094, "step": 76480 }, { "grad_norm": 0.27469462156295776, "learning_rate": 5.256682837721627e-07, "loss": 0.0155, "step": 76490 }, { "grad_norm": 0.0756952166557312, "learning_rate": 5.226833609435156e-07, "loss": 0.0098, "step": 76500 }, { "grad_norm": 0.3846120238304138, "learning_rate": 5.19706892435079e-07, "loss": 0.0131, "step": 76510 }, { "grad_norm": 0.1302325576543808, "learning_rate": 5.167388787554406e-07, "loss": 0.0095, "step": 76520 }, { "grad_norm": 0.13425561785697937, "learning_rate": 5.137793204117614e-07, "loss": 0.0104, "step": 76530 }, { "grad_norm": 0.34682267904281616, "learning_rate": 5.108282179097479e-07, "loss": 0.0127, "step": 76540 }, { "grad_norm": 0.2244754582643509, "learning_rate": 5.078855717536523e-07, "loss": 0.0108, "step": 76550 }, { "grad_norm": 0.30973732471466064, "learning_rate": 5.049513824463059e-07, "loss": 0.0085, "step": 76560 }, { "grad_norm": 0.3742080628871918, "learning_rate": 5.020256504890742e-07, "loss": 0.0133, "step": 76570 }, { "grad_norm": 0.24890755116939545, "learning_rate": 4.991083763818849e-07, "loss": 0.0132, "step": 76580 }, { "grad_norm": 0.4306381046772003, "learning_rate": 4.961995606232228e-07, "loss": 0.0136, "step": 76590 }, { "grad_norm": 0.1407700777053833, "learning_rate": 4.932992037101236e-07, "loss": 0.0097, "step": 76600 }, { "grad_norm": 0.3884439468383789, "learning_rate": 4.904073061381798e-07, "loss": 0.0163, "step": 76610 }, { "grad_norm": 0.1339663565158844, "learning_rate": 4.875238684015349e-07, "loss": 0.0103, "step": 76620 }, { "grad_norm": 0.2988234758377075, "learning_rate": 4.846488909928948e-07, "loss": 0.0092, "step": 76630 }, { "grad_norm": 0.13960817456245422, "learning_rate": 4.81782374403511e-07, "loss": 0.0144, "step": 76640 }, { "grad_norm": 0.23750491440296173, "learning_rate": 4.789243191231918e-07, "loss": 0.0088, "step": 76650 }, { "grad_norm": 0.5911957621574402, "learning_rate": 4.76074725640302e-07, "loss": 0.017, "step": 76660 }, { "grad_norm": 0.171881303191185, "learning_rate": 4.732335944417632e-07, "loss": 0.0166, "step": 76670 }, { "grad_norm": 0.21429435908794403, "learning_rate": 4.704009260130371e-07, "loss": 0.0166, "step": 76680 }, { "grad_norm": 0.22432464361190796, "learning_rate": 4.675767208381587e-07, "loss": 0.0093, "step": 76690 }, { "grad_norm": 0.19108952581882477, "learning_rate": 4.647609793997032e-07, "loss": 0.0098, "step": 76700 }, { "grad_norm": 0.24810396134853363, "learning_rate": 4.619537021788023e-07, "loss": 0.011, "step": 76710 }, { "grad_norm": 0.2333662360906601, "learning_rate": 4.5915488965515005e-07, "loss": 0.0118, "step": 76720 }, { "grad_norm": 0.2197331190109253, "learning_rate": 4.563645423069807e-07, "loss": 0.012, "step": 76730 }, { "grad_norm": 0.31248727440834045, "learning_rate": 4.535826606110849e-07, "loss": 0.0123, "step": 76740 }, { "grad_norm": 0.37254494428634644, "learning_rate": 4.508092450428214e-07, "loss": 0.0147, "step": 76750 }, { "grad_norm": 0.14806826412677765, "learning_rate": 4.480442960760778e-07, "loss": 0.011, "step": 76760 }, { "grad_norm": 0.1740458756685257, "learning_rate": 4.4528781418332053e-07, "loss": 0.0095, "step": 76770 }, { "grad_norm": 0.2966366112232208, "learning_rate": 4.4253979983555073e-07, "loss": 0.0148, "step": 76780 }, { "grad_norm": 0.13971365988254547, "learning_rate": 4.3980025350233154e-07, "loss": 0.0091, "step": 76790 }, { "grad_norm": 0.2178233563899994, "learning_rate": 4.370691756517664e-07, "loss": 0.012, "step": 76800 }, { "grad_norm": 0.23372216522693634, "learning_rate": 4.3434656675053753e-07, "loss": 0.0136, "step": 76810 }, { "grad_norm": 0.2466305047273636, "learning_rate": 4.3163242726385613e-07, "loss": 0.0124, "step": 76820 }, { "grad_norm": 0.1924758404493332, "learning_rate": 4.289267576554956e-07, "loss": 0.0153, "step": 76830 }, { "grad_norm": 0.23988081514835358, "learning_rate": 4.262295583877807e-07, "loss": 0.0118, "step": 76840 }, { "grad_norm": 0.07281787693500519, "learning_rate": 4.235408299215815e-07, "loss": 0.0171, "step": 76850 }, { "grad_norm": 0.2428751140832901, "learning_rate": 4.2086057271634173e-07, "loss": 0.0122, "step": 76860 }, { "grad_norm": 0.2559163272380829, "learning_rate": 4.181887872300394e-07, "loss": 0.015, "step": 76870 }, { "grad_norm": 0.35835883021354675, "learning_rate": 4.155254739192038e-07, "loss": 0.0111, "step": 76880 }, { "grad_norm": 0.19503037631511688, "learning_rate": 4.128706332389265e-07, "loss": 0.013, "step": 76890 }, { "grad_norm": 0.40251269936561584, "learning_rate": 4.1022426564284453e-07, "loss": 0.0149, "step": 76900 }, { "grad_norm": 0.24134580790996552, "learning_rate": 4.075863715831574e-07, "loss": 0.0123, "step": 76910 }, { "grad_norm": 0.27614372968673706, "learning_rate": 4.0495695151059887e-07, "loss": 0.014, "step": 76920 }, { "grad_norm": 0.9133927226066589, "learning_rate": 4.023360058744763e-07, "loss": 0.0103, "step": 76930 }, { "grad_norm": 0.20282316207885742, "learning_rate": 3.997235351226314e-07, "loss": 0.0091, "step": 76940 }, { "grad_norm": 0.25259101390838623, "learning_rate": 3.971195397014571e-07, "loss": 0.0139, "step": 76950 }, { "grad_norm": 0.35626882314682007, "learning_rate": 3.9452402005591417e-07, "loss": 0.0152, "step": 76960 }, { "grad_norm": 0.1858721822500229, "learning_rate": 3.9193697662950334e-07, "loss": 0.0109, "step": 76970 }, { "grad_norm": 0.624013364315033, "learning_rate": 3.893584098642822e-07, "loss": 0.0111, "step": 76980 }, { "grad_norm": 0.17135940492153168, "learning_rate": 3.867883202008538e-07, "loss": 0.0161, "step": 76990 }, { "grad_norm": 0.21929602324962616, "learning_rate": 3.842267080783779e-07, "loss": 0.0092, "step": 77000 }, { "grad_norm": 0.12797784805297852, "learning_rate": 3.816735739345656e-07, "loss": 0.0104, "step": 77010 }, { "grad_norm": 0.11210527271032333, "learning_rate": 3.791289182056679e-07, "loss": 0.0093, "step": 77020 }, { "grad_norm": 0.22211068868637085, "learning_rate": 3.7659274132650913e-07, "loss": 0.0122, "step": 77030 }, { "grad_norm": 0.2236514389514923, "learning_rate": 3.7406504373044826e-07, "loss": 0.0099, "step": 77040 }, { "grad_norm": 0.1688978523015976, "learning_rate": 3.7154582584939533e-07, "loss": 0.01, "step": 77050 }, { "grad_norm": 0.2267017811536789, "learning_rate": 3.690350881138227e-07, "loss": 0.0114, "step": 77060 }, { "grad_norm": 0.3475680947303772, "learning_rate": 3.665328309527427e-07, "loss": 0.0124, "step": 77070 }, { "grad_norm": 0.29090777039527893, "learning_rate": 3.640390547937245e-07, "loss": 0.0145, "step": 77080 }, { "grad_norm": 0.4471585154533386, "learning_rate": 3.615537600628882e-07, "loss": 0.0158, "step": 77090 }, { "grad_norm": 0.32405418157577515, "learning_rate": 3.5907694718489427e-07, "loss": 0.0124, "step": 77100 }, { "grad_norm": 0.28107163310050964, "learning_rate": 3.566086165829707e-07, "loss": 0.0169, "step": 77110 }, { "grad_norm": 0.4875236749649048, "learning_rate": 3.5414876867888024e-07, "loss": 0.0128, "step": 77120 }, { "grad_norm": 0.29012560844421387, "learning_rate": 3.5169740389295326e-07, "loss": 0.0163, "step": 77130 }, { "grad_norm": 0.1918303519487381, "learning_rate": 3.4925452264405466e-07, "loss": 0.0123, "step": 77140 }, { "grad_norm": 0.14974980056285858, "learning_rate": 3.468201253496062e-07, "loss": 0.0107, "step": 77150 }, { "grad_norm": 0.2852492928504944, "learning_rate": 3.443942124255861e-07, "loss": 0.0139, "step": 77160 }, { "grad_norm": 0.3218659460544586, "learning_rate": 3.4197678428650183e-07, "loss": 0.0143, "step": 77170 }, { "grad_norm": 0.17543280124664307, "learning_rate": 3.3956784134544504e-07, "loss": 0.0126, "step": 77180 }, { "grad_norm": 0.2240850180387497, "learning_rate": 3.3716738401402547e-07, "loss": 0.0116, "step": 77190 }, { "grad_norm": 0.30270352959632874, "learning_rate": 3.3477541270241495e-07, "loss": 0.0099, "step": 77200 }, { "grad_norm": 0.23028500378131866, "learning_rate": 3.3239192781934215e-07, "loss": 0.0091, "step": 77210 }, { "grad_norm": 0.15723493695259094, "learning_rate": 3.3001692977207563e-07, "loss": 0.01, "step": 77220 }, { "grad_norm": 0.11145713180303574, "learning_rate": 3.276504189664409e-07, "loss": 0.0099, "step": 77230 }, { "grad_norm": 0.18950894474983215, "learning_rate": 3.252923958068088e-07, "loss": 0.0135, "step": 77240 }, { "grad_norm": 0.2888704538345337, "learning_rate": 3.2294286069609046e-07, "loss": 0.0113, "step": 77250 }, { "grad_norm": 0.24769873917102814, "learning_rate": 3.2060181403577583e-07, "loss": 0.0112, "step": 77260 }, { "grad_norm": 0.29781922698020935, "learning_rate": 3.1826925622587823e-07, "loss": 0.0108, "step": 77270 }, { "grad_norm": 0.1477038562297821, "learning_rate": 3.159451876649622e-07, "loss": 0.0128, "step": 77280 }, { "grad_norm": 0.18908846378326416, "learning_rate": 3.1362960875015444e-07, "loss": 0.0087, "step": 77290 }, { "grad_norm": 0.3658246695995331, "learning_rate": 3.1132251987711637e-07, "loss": 0.0124, "step": 77300 }, { "grad_norm": 0.25508180260658264, "learning_rate": 3.0902392144007695e-07, "loss": 0.0102, "step": 77310 }, { "grad_norm": 0.19892370700836182, "learning_rate": 3.0673381383179436e-07, "loss": 0.0122, "step": 77320 }, { "grad_norm": 0.2961159944534302, "learning_rate": 3.0445219744358875e-07, "loss": 0.0118, "step": 77330 }, { "grad_norm": 0.15959738194942474, "learning_rate": 3.021790726653262e-07, "loss": 0.0162, "step": 77340 }, { "grad_norm": 0.3883723020553589, "learning_rate": 2.9991443988542366e-07, "loss": 0.0144, "step": 77350 }, { "grad_norm": 0.4867755174636841, "learning_rate": 2.976582994908439e-07, "loss": 0.0103, "step": 77360 }, { "grad_norm": 0.14666812121868134, "learning_rate": 2.954106518671007e-07, "loss": 0.0106, "step": 77370 }, { "grad_norm": 0.26820915937423706, "learning_rate": 2.9317149739825356e-07, "loss": 0.0134, "step": 77380 }, { "grad_norm": 0.5264991521835327, "learning_rate": 2.909408364669075e-07, "loss": 0.016, "step": 77390 }, { "grad_norm": 0.32790639996528625, "learning_rate": 2.887186694542299e-07, "loss": 0.0126, "step": 77400 }, { "grad_norm": 0.16236792504787445, "learning_rate": 2.865049967399225e-07, "loss": 0.0104, "step": 77410 }, { "grad_norm": 0.1475883573293686, "learning_rate": 2.842998187022439e-07, "loss": 0.0113, "step": 77420 }, { "grad_norm": 0.19974149763584137, "learning_rate": 2.8210313571800374e-07, "loss": 0.0126, "step": 77430 }, { "grad_norm": 0.25089022517204285, "learning_rate": 2.7991494816255184e-07, "loss": 0.0099, "step": 77440 }, { "grad_norm": 0.27580031752586365, "learning_rate": 2.777352564097779e-07, "loss": 0.0131, "step": 77450 }, { "grad_norm": 0.22275878489017487, "learning_rate": 2.755640608321508e-07, "loss": 0.0117, "step": 77460 }, { "grad_norm": 0.3408229649066925, "learning_rate": 2.73401361800657e-07, "loss": 0.0119, "step": 77470 }, { "grad_norm": 0.3315279185771942, "learning_rate": 2.7124715968484537e-07, "loss": 0.0125, "step": 77480 }, { "grad_norm": 0.28813618421554565, "learning_rate": 2.691014548528104e-07, "loss": 0.0148, "step": 77490 }, { "grad_norm": 0.20033566653728485, "learning_rate": 2.669642476711864e-07, "loss": 0.0148, "step": 77500 }, { "grad_norm": 0.3889892101287842, "learning_rate": 2.648355385051815e-07, "loss": 0.0125, "step": 77510 }, { "grad_norm": 0.4765082001686096, "learning_rate": 2.627153277185157e-07, "loss": 0.0124, "step": 77520 }, { "grad_norm": 0.14202582836151123, "learning_rate": 2.606036156734881e-07, "loss": 0.013, "step": 77530 }, { "grad_norm": 0.1249777302145958, "learning_rate": 2.5850040273092127e-07, "loss": 0.0124, "step": 77540 }, { "grad_norm": 0.2100343406200409, "learning_rate": 2.5640568925020536e-07, "loss": 0.0214, "step": 77550 }, { "grad_norm": 0.19124844670295715, "learning_rate": 2.5431947558927083e-07, "loss": 0.0115, "step": 77560 }, { "grad_norm": 0.2099418342113495, "learning_rate": 2.522417621045825e-07, "loss": 0.0099, "step": 77570 }, { "grad_norm": 0.384430468082428, "learning_rate": 2.501725491511786e-07, "loss": 0.0148, "step": 77580 }, { "grad_norm": 0.3278273046016693, "learning_rate": 2.481118370826263e-07, "loss": 0.0128, "step": 77590 }, { "grad_norm": 0.19128596782684326, "learning_rate": 2.4605962625104393e-07, "loss": 0.0107, "step": 77600 }, { "grad_norm": 0.3724776804447174, "learning_rate": 2.440159170070955e-07, "loss": 0.0149, "step": 77610 }, { "grad_norm": 0.21173036098480225, "learning_rate": 2.41980709699996e-07, "loss": 0.0142, "step": 77620 }, { "grad_norm": 0.16542468965053558, "learning_rate": 2.3995400467751174e-07, "loss": 0.0126, "step": 77630 }, { "grad_norm": 0.18782848119735718, "learning_rate": 2.3793580228594902e-07, "loss": 0.0158, "step": 77640 }, { "grad_norm": 0.2593756914138794, "learning_rate": 2.3592610287015982e-07, "loss": 0.0097, "step": 77650 }, { "grad_norm": 0.1660415679216385, "learning_rate": 2.3392490677354718e-07, "loss": 0.0102, "step": 77660 }, { "grad_norm": 0.2611472010612488, "learning_rate": 2.3193221433806533e-07, "loss": 0.0088, "step": 77670 }, { "grad_norm": 0.22931477427482605, "learning_rate": 2.299480259042086e-07, "loss": 0.0122, "step": 77680 }, { "grad_norm": 0.2864161729812622, "learning_rate": 2.2797234181102244e-07, "loss": 0.0146, "step": 77690 }, { "grad_norm": 0.20285479724407196, "learning_rate": 2.2600516239609238e-07, "loss": 0.0104, "step": 77700 }, { "grad_norm": 0.22664669156074524, "learning_rate": 2.240464879955606e-07, "loss": 0.0109, "step": 77710 }, { "grad_norm": 0.2841378450393677, "learning_rate": 2.2209631894410387e-07, "loss": 0.0126, "step": 77720 }, { "grad_norm": 0.24363800883293152, "learning_rate": 2.2015465557496117e-07, "loss": 0.0124, "step": 77730 }, { "grad_norm": 0.2819400727748871, "learning_rate": 2.1822149821990602e-07, "loss": 0.0139, "step": 77740 }, { "grad_norm": 0.3890259563922882, "learning_rate": 2.1629684720926303e-07, "loss": 0.0162, "step": 77750 }, { "grad_norm": 0.20207896828651428, "learning_rate": 2.1438070287189693e-07, "loss": 0.0103, "step": 77760 }, { "grad_norm": 0.17034190893173218, "learning_rate": 2.1247306553523472e-07, "loss": 0.0129, "step": 77770 }, { "grad_norm": 0.23436106741428375, "learning_rate": 2.105739355252323e-07, "loss": 0.0106, "step": 77780 }, { "grad_norm": 0.1539381593465805, "learning_rate": 2.0868331316639678e-07, "loss": 0.01, "step": 77790 }, { "grad_norm": 0.5497621297836304, "learning_rate": 2.0680119878179193e-07, "loss": 0.0114, "step": 77800 }, { "grad_norm": 0.32845932245254517, "learning_rate": 2.0492759269301054e-07, "loss": 0.0082, "step": 77810 }, { "grad_norm": 0.48596659302711487, "learning_rate": 2.0306249522021315e-07, "loss": 0.0117, "step": 77820 }, { "grad_norm": 0.20581094920635223, "learning_rate": 2.0120590668207816e-07, "loss": 0.0101, "step": 77830 }, { "grad_norm": 0.16584385931491852, "learning_rate": 1.993578273958574e-07, "loss": 0.013, "step": 77840 }, { "grad_norm": 0.16577933728694916, "learning_rate": 1.975182576773371e-07, "loss": 0.0087, "step": 77850 }, { "grad_norm": 0.5692639946937561, "learning_rate": 1.9568719784083812e-07, "loss": 0.0148, "step": 77860 }, { "grad_norm": 0.21993637084960938, "learning_rate": 1.9386464819924898e-07, "loss": 0.0127, "step": 77870 }, { "grad_norm": 0.12132833153009415, "learning_rate": 1.9205060906399285e-07, "loss": 0.0097, "step": 77880 }, { "grad_norm": 0.6842681169509888, "learning_rate": 1.9024508074503845e-07, "loss": 0.0138, "step": 77890 }, { "grad_norm": 0.3449883759021759, "learning_rate": 1.8844806355089452e-07, "loss": 0.0146, "step": 77900 }, { "grad_norm": 0.4711892306804657, "learning_rate": 1.8665955778863208e-07, "loss": 0.0137, "step": 77910 }, { "grad_norm": 0.2464582622051239, "learning_rate": 1.848795637638512e-07, "loss": 0.0109, "step": 77920 }, { "grad_norm": 0.2547699809074402, "learning_rate": 1.831080817807085e-07, "loss": 0.0137, "step": 77930 }, { "grad_norm": 0.3210512697696686, "learning_rate": 1.813451121418952e-07, "loss": 0.0213, "step": 77940 }, { "grad_norm": 0.4330842196941376, "learning_rate": 1.795906551486648e-07, "loss": 0.0109, "step": 77950 }, { "grad_norm": 0.38193556666374207, "learning_rate": 1.7784471110079414e-07, "loss": 0.0135, "step": 77960 }, { "grad_norm": 0.2734374701976776, "learning_rate": 1.7610728029662793e-07, "loss": 0.0118, "step": 77970 }, { "grad_norm": 0.22104176878929138, "learning_rate": 1.7437836303303979e-07, "loss": 0.0117, "step": 77980 }, { "grad_norm": 0.15491797029972076, "learning_rate": 1.726579596054545e-07, "loss": 0.0084, "step": 77990 }, { "grad_norm": 0.3145291209220886, "learning_rate": 1.7094607030784803e-07, "loss": 0.0119, "step": 78000 }, { "grad_norm": 0.1362445056438446, "learning_rate": 1.6924269543272532e-07, "loss": 0.0107, "step": 78010 }, { "grad_norm": 0.4925210177898407, "learning_rate": 1.6754783527115348e-07, "loss": 0.0144, "step": 78020 }, { "grad_norm": 0.11367945373058319, "learning_rate": 1.6586149011273422e-07, "loss": 0.0128, "step": 78030 }, { "grad_norm": 0.3693835735321045, "learning_rate": 1.6418366024562038e-07, "loss": 0.0179, "step": 78040 }, { "grad_norm": 0.24226249754428864, "learning_rate": 1.6251434595651037e-07, "loss": 0.0129, "step": 78050 }, { "grad_norm": 0.20612478256225586, "learning_rate": 1.608535475306372e-07, "loss": 0.0129, "step": 78060 }, { "grad_norm": 0.19465240836143494, "learning_rate": 1.5920126525179048e-07, "loss": 0.008, "step": 78070 }, { "grad_norm": 0.14304345846176147, "learning_rate": 1.5755749940229435e-07, "loss": 0.0089, "step": 78080 }, { "grad_norm": 0.2565038502216339, "learning_rate": 1.559222502630353e-07, "loss": 0.0116, "step": 78090 }, { "grad_norm": 0.3400079905986786, "learning_rate": 1.5429551811341757e-07, "loss": 0.0148, "step": 78100 }, { "grad_norm": 0.6531775593757629, "learning_rate": 1.5267730323141882e-07, "loss": 0.0118, "step": 78110 }, { "grad_norm": 0.19629256427288055, "learning_rate": 1.5106760589353454e-07, "loss": 0.0104, "step": 78120 }, { "grad_norm": 0.3058010935783386, "learning_rate": 1.4946642637483355e-07, "loss": 0.0106, "step": 78130 }, { "grad_norm": 0.303548663854599, "learning_rate": 1.4787376494889703e-07, "loss": 0.0147, "step": 78140 }, { "grad_norm": 0.1630536913871765, "learning_rate": 1.4628962188787955e-07, "loss": 0.0119, "step": 78150 }, { "grad_norm": 0.29853126406669617, "learning_rate": 1.4471399746247006e-07, "loss": 0.0134, "step": 78160 }, { "grad_norm": 0.20687302947044373, "learning_rate": 1.4314689194188103e-07, "loss": 0.0123, "step": 78170 }, { "grad_norm": 0.5948399305343628, "learning_rate": 1.4158830559390933e-07, "loss": 0.0105, "step": 78180 }, { "grad_norm": 0.2049078494310379, "learning_rate": 1.4003823868486422e-07, "loss": 0.0232, "step": 78190 }, { "grad_norm": 0.333553671836853, "learning_rate": 1.3849669147960598e-07, "loss": 0.0166, "step": 78200 }, { "grad_norm": 0.20126719772815704, "learning_rate": 1.3696366424155726e-07, "loss": 0.0123, "step": 78210 }, { "grad_norm": 0.4643575847148895, "learning_rate": 1.35439157232653e-07, "loss": 0.0195, "step": 78220 }, { "grad_norm": 0.46895086765289307, "learning_rate": 1.3392317071340144e-07, "loss": 0.012, "step": 78230 }, { "grad_norm": 0.20007772743701935, "learning_rate": 1.3241570494283984e-07, "loss": 0.0111, "step": 78240 }, { "grad_norm": 0.20848402380943298, "learning_rate": 1.3091676017855103e-07, "loss": 0.0172, "step": 78250 }, { "grad_norm": 0.2686605751514435, "learning_rate": 1.2942633667666904e-07, "loss": 0.0096, "step": 78260 }, { "grad_norm": 0.265287309885025, "learning_rate": 1.2794443469185679e-07, "loss": 0.0168, "step": 78270 }, { "grad_norm": 0.5385063290596008, "learning_rate": 1.2647105447734508e-07, "loss": 0.0104, "step": 78280 }, { "grad_norm": 0.2948491871356964, "learning_rate": 1.2500619628488254e-07, "loss": 0.0116, "step": 78290 }, { "grad_norm": 0.3391103446483612, "learning_rate": 1.2354986036477446e-07, "loss": 0.0158, "step": 78300 }, { "grad_norm": 0.5977433323860168, "learning_rate": 1.221020469658718e-07, "loss": 0.0148, "step": 78310 }, { "grad_norm": 0.20926536619663239, "learning_rate": 1.2066275633556556e-07, "loss": 0.0134, "step": 78320 }, { "grad_norm": 0.315233051776886, "learning_rate": 1.192319887197979e-07, "loss": 0.0123, "step": 78330 }, { "grad_norm": 0.22263282537460327, "learning_rate": 1.1780974436303438e-07, "loss": 0.0098, "step": 78340 }, { "grad_norm": 0.17578372359275818, "learning_rate": 1.1639602350830836e-07, "loss": 0.0121, "step": 78350 }, { "grad_norm": 0.3176206350326538, "learning_rate": 1.1499082639718217e-07, "loss": 0.0097, "step": 78360 }, { "grad_norm": 0.1712176650762558, "learning_rate": 1.1359415326976374e-07, "loss": 0.0112, "step": 78370 }, { "grad_norm": 0.16062532365322113, "learning_rate": 1.1220600436470663e-07, "loss": 0.0126, "step": 78380 }, { "grad_norm": 0.25666528940200806, "learning_rate": 1.1082637991920996e-07, "loss": 0.0147, "step": 78390 }, { "grad_norm": 0.1762809455394745, "learning_rate": 1.0945528016901851e-07, "loss": 0.0121, "step": 78400 }, { "grad_norm": 0.34564563632011414, "learning_rate": 1.0809270534840599e-07, "loss": 0.0098, "step": 78410 }, { "grad_norm": 0.147648423910141, "learning_rate": 1.067386556902028e-07, "loss": 0.0097, "step": 78420 }, { "grad_norm": 0.1340702623128891, "learning_rate": 1.05393131425785e-07, "loss": 0.0068, "step": 78430 }, { "grad_norm": 0.360953688621521, "learning_rate": 1.0405613278505199e-07, "loss": 0.0113, "step": 78440 }, { "grad_norm": 0.3311816155910492, "learning_rate": 1.027276599964766e-07, "loss": 0.0123, "step": 78450 }, { "grad_norm": 0.23376509547233582, "learning_rate": 1.0140771328704391e-07, "loss": 0.0107, "step": 78460 }, { "grad_norm": 0.2281157672405243, "learning_rate": 1.0009629288231237e-07, "loss": 0.0077, "step": 78470 }, { "grad_norm": 0.4044528901576996, "learning_rate": 9.879339900635276e-08, "loss": 0.0142, "step": 78480 }, { "grad_norm": 0.3417785167694092, "learning_rate": 9.74990318817981e-08, "loss": 0.0125, "step": 78490 }, { "grad_norm": 0.33296796679496765, "learning_rate": 9.621319172982701e-08, "loss": 0.0119, "step": 78500 }, { "grad_norm": 0.20443716645240784, "learning_rate": 9.493587877015264e-08, "loss": 0.0097, "step": 78510 }, { "grad_norm": 0.1805301010608673, "learning_rate": 9.366709322102263e-08, "loss": 0.0122, "step": 78520 }, { "grad_norm": 0.2603277862071991, "learning_rate": 9.240683529924688e-08, "loss": 0.0125, "step": 78530 }, { "grad_norm": 0.3115304708480835, "learning_rate": 9.115510522016979e-08, "loss": 0.0124, "step": 78540 }, { "grad_norm": 0.28228065371513367, "learning_rate": 8.991190319767583e-08, "loss": 0.01, "step": 78550 }, { "grad_norm": 0.15507353842258453, "learning_rate": 8.867722944419509e-08, "loss": 0.0111, "step": 78560 }, { "grad_norm": 0.23667660355567932, "learning_rate": 8.745108417069214e-08, "loss": 0.0104, "step": 78570 }, { "grad_norm": 0.23157531023025513, "learning_rate": 8.623346758669381e-08, "loss": 0.0125, "step": 78580 }, { "grad_norm": 0.3504512310028076, "learning_rate": 8.502437990025037e-08, "loss": 0.0116, "step": 78590 }, { "grad_norm": 0.20291896164417267, "learning_rate": 8.382382131795764e-08, "loss": 0.0126, "step": 78600 }, { "grad_norm": 0.18911047279834747, "learning_rate": 8.263179204496818e-08, "loss": 0.0129, "step": 78610 }, { "grad_norm": 0.3719596862792969, "learning_rate": 8.144829228496354e-08, "loss": 0.012, "step": 78620 }, { "grad_norm": 0.3651370704174042, "learning_rate": 8.027332224016526e-08, "loss": 0.0141, "step": 78630 }, { "grad_norm": 0.2750798463821411, "learning_rate": 7.910688211135164e-08, "loss": 0.012, "step": 78640 }, { "grad_norm": 0.25978294014930725, "learning_rate": 7.794897209783546e-08, "loss": 0.0118, "step": 78650 }, { "grad_norm": 0.3216742277145386, "learning_rate": 7.679959239746403e-08, "loss": 0.0157, "step": 78660 }, { "grad_norm": 0.16909517347812653, "learning_rate": 7.565874320664135e-08, "loss": 0.0129, "step": 78670 }, { "grad_norm": 0.6054089665412903, "learning_rate": 7.45264247203059e-08, "loss": 0.0148, "step": 78680 }, { "grad_norm": 0.19592930376529694, "learning_rate": 7.340263713194184e-08, "loss": 0.0113, "step": 78690 }, { "grad_norm": 0.2082356959581375, "learning_rate": 7.228738063356777e-08, "loss": 0.0082, "step": 78700 }, { "grad_norm": 0.2597024738788605, "learning_rate": 7.118065541575903e-08, "loss": 0.0115, "step": 78710 }, { "grad_norm": 0.25240838527679443, "learning_rate": 7.008246166761435e-08, "loss": 0.012, "step": 78720 }, { "grad_norm": 0.24441242218017578, "learning_rate": 6.899279957679472e-08, "loss": 0.0097, "step": 78730 }, { "grad_norm": 0.1998971849679947, "learning_rate": 6.791166932949011e-08, "loss": 0.0116, "step": 78740 }, { "grad_norm": 0.3204055428504944, "learning_rate": 6.683907111043608e-08, "loss": 0.0094, "step": 78750 }, { "grad_norm": 0.18598710000514984, "learning_rate": 6.577500510290824e-08, "loss": 0.0115, "step": 78760 }, { "grad_norm": 0.585691511631012, "learning_rate": 6.47194714887278e-08, "loss": 0.02, "step": 78770 }, { "grad_norm": 0.2769066393375397, "learning_rate": 6.367247044825608e-08, "loss": 0.0143, "step": 78780 }, { "grad_norm": 0.35713180899620056, "learning_rate": 6.263400216039994e-08, "loss": 0.0096, "step": 78790 }, { "grad_norm": 0.7909667491912842, "learning_rate": 6.160406680260078e-08, "loss": 0.0212, "step": 78800 }, { "grad_norm": 0.24386487901210785, "learning_rate": 6.058266455084561e-08, "loss": 0.0115, "step": 78810 }, { "grad_norm": 0.47440552711486816, "learning_rate": 5.956979557967257e-08, "loss": 0.0116, "step": 78820 }, { "grad_norm": 0.11717603355646133, "learning_rate": 5.856546006214325e-08, "loss": 0.0116, "step": 78830 }, { "grad_norm": 0.20669089257717133, "learning_rate": 5.7569658169881466e-08, "loss": 0.0106, "step": 78840 }, { "grad_norm": 0.15684646368026733, "learning_rate": 5.658239007303445e-08, "loss": 0.015, "step": 78850 }, { "grad_norm": 0.32167795300483704, "learning_rate": 5.560365594030059e-08, "loss": 0.0148, "step": 78860 }, { "grad_norm": 0.1408850997686386, "learning_rate": 5.463345593891833e-08, "loss": 0.0086, "step": 78870 }, { "grad_norm": 0.2252030074596405, "learning_rate": 5.367179023467173e-08, "loss": 0.0132, "step": 78880 }, { "grad_norm": 0.1125437393784523, "learning_rate": 5.271865899187378e-08, "loss": 0.0108, "step": 78890 }, { "grad_norm": 0.22164717316627502, "learning_rate": 5.177406237340532e-08, "loss": 0.0181, "step": 78900 }, { "grad_norm": 0.22551944851875305, "learning_rate": 5.083800054065946e-08, "loss": 0.0105, "step": 78910 }, { "grad_norm": 0.22472742199897766, "learning_rate": 4.991047365358603e-08, "loss": 0.0116, "step": 78920 }, { "grad_norm": 0.520626425743103, "learning_rate": 4.899148187067493e-08, "loss": 0.0148, "step": 78930 }, { "grad_norm": 0.18396644294261932, "learning_rate": 4.808102534895609e-08, "loss": 0.0085, "step": 78940 }, { "grad_norm": 0.3101074695587158, "learning_rate": 4.717910424400507e-08, "loss": 0.0111, "step": 78950 }, { "grad_norm": 0.31034055352211, "learning_rate": 4.628571870993193e-08, "loss": 0.0146, "step": 78960 }, { "grad_norm": 0.20498166978359222, "learning_rate": 4.5400868899392324e-08, "loss": 0.0135, "step": 78970 }, { "grad_norm": 0.4403424859046936, "learning_rate": 4.4524554963576436e-08, "loss": 0.0156, "step": 78980 }, { "grad_norm": 0.35967564582824707, "learning_rate": 4.36567770522367e-08, "loss": 0.0142, "step": 78990 }, { "grad_norm": 0.18196523189544678, "learning_rate": 4.279753531364339e-08, "loss": 0.0087, "step": 79000 }, { "grad_norm": 0.5341916680335999, "learning_rate": 4.1946829894617955e-08, "loss": 0.0136, "step": 79010 }, { "grad_norm": 0.27783387899398804, "learning_rate": 4.11046609405219e-08, "loss": 0.0164, "step": 79020 }, { "grad_norm": 0.2608017325401306, "learning_rate": 4.027102859526233e-08, "loss": 0.0065, "step": 79030 }, { "grad_norm": 0.39667966961860657, "learning_rate": 3.9445933001280856e-08, "loss": 0.0116, "step": 79040 }, { "grad_norm": 0.21625179052352905, "learning_rate": 3.8629374299564704e-08, "loss": 0.0122, "step": 79050 }, { "grad_norm": 1.4173263311386108, "learning_rate": 3.78213526296467e-08, "loss": 0.0156, "step": 79060 }, { "grad_norm": 0.22426508367061615, "learning_rate": 3.702186812958308e-08, "loss": 0.0146, "step": 79070 }, { "grad_norm": 0.23812180757522583, "learning_rate": 3.623092093599789e-08, "loss": 0.0104, "step": 79080 }, { "grad_norm": 0.19042794406414032, "learning_rate": 3.544851118403303e-08, "loss": 0.0105, "step": 79090 }, { "grad_norm": 0.6851698160171509, "learning_rate": 3.46746390073871e-08, "loss": 0.0149, "step": 79100 }, { "grad_norm": 0.19758866727352142, "learning_rate": 3.390930453829322e-08, "loss": 0.0194, "step": 79110 }, { "grad_norm": 0.08097810298204422, "learning_rate": 3.3152507907519005e-08, "loss": 0.0099, "step": 79120 }, { "grad_norm": 0.39511919021606445, "learning_rate": 3.240424924438323e-08, "loss": 0.0123, "step": 79130 }, { "grad_norm": 0.44778627157211304, "learning_rate": 3.1664528676750295e-08, "loss": 0.0122, "step": 79140 }, { "grad_norm": 0.25386252999305725, "learning_rate": 3.093334633100797e-08, "loss": 0.0153, "step": 79150 }, { "grad_norm": 0.2704463601112366, "learning_rate": 3.021070233210077e-08, "loss": 0.0128, "step": 79160 }, { "grad_norm": 0.4272395372390747, "learning_rate": 2.9496596803507692e-08, "loss": 0.0088, "step": 79170 }, { "grad_norm": 0.1723255068063736, "learning_rate": 2.879102986725335e-08, "loss": 0.0101, "step": 79180 }, { "grad_norm": 0.21639679372310638, "learning_rate": 2.809400164389131e-08, "loss": 0.0113, "step": 79190 }, { "grad_norm": 0.26480719447135925, "learning_rate": 2.740551225253185e-08, "loss": 0.0142, "step": 79200 }, { "grad_norm": 0.2905847430229187, "learning_rate": 2.6725561810819754e-08, "loss": 0.0111, "step": 79210 }, { "grad_norm": 0.2462221086025238, "learning_rate": 2.605415043493431e-08, "loss": 0.0125, "step": 79220 }, { "grad_norm": 0.43223896622657776, "learning_rate": 2.5391278239605965e-08, "loss": 0.0146, "step": 79230 }, { "grad_norm": 0.18767738342285156, "learning_rate": 2.4736945338094118e-08, "loss": 0.0111, "step": 79240 }, { "grad_norm": 0.13740210235118866, "learning_rate": 2.4091151842214887e-08, "loss": 0.0091, "step": 79250 }, { "grad_norm": 0.3801391124725342, "learning_rate": 2.3453897862318884e-08, "loss": 0.0164, "step": 79260 }, { "grad_norm": 0.2003704458475113, "learning_rate": 2.2825183507285686e-08, "loss": 0.0133, "step": 79270 }, { "grad_norm": 0.2907314598560333, "learning_rate": 2.220500888455157e-08, "loss": 0.0177, "step": 79280 }, { "grad_norm": 0.28669437766075134, "learning_rate": 2.1593374100081777e-08, "loss": 0.0153, "step": 79290 }, { "grad_norm": 0.17114830017089844, "learning_rate": 2.09902792583927e-08, "loss": 0.0173, "step": 79300 }, { "grad_norm": 0.16699399054050446, "learning_rate": 2.0395724462540788e-08, "loss": 0.0123, "step": 79310 }, { "grad_norm": 0.26966404914855957, "learning_rate": 1.9809709814111453e-08, "loss": 0.0107, "step": 79320 }, { "grad_norm": 0.14341531693935394, "learning_rate": 1.923223541324126e-08, "loss": 0.018, "step": 79330 }, { "grad_norm": 0.16234971582889557, "learning_rate": 1.8663301358606833e-08, "loss": 0.0146, "step": 79340 }, { "grad_norm": 0.31713223457336426, "learning_rate": 1.8102907747419295e-08, "loss": 0.0141, "step": 79350 }, { "grad_norm": 0.25834307074546814, "learning_rate": 1.7551054675435385e-08, "loss": 0.0101, "step": 79360 }, { "grad_norm": 0.42739325761795044, "learning_rate": 1.7007742236957447e-08, "loss": 0.0128, "step": 79370 }, { "grad_norm": 0.28649085760116577, "learning_rate": 1.647297052481678e-08, "loss": 0.0127, "step": 79380 }, { "grad_norm": 0.18770451843738556, "learning_rate": 1.5946739630390285e-08, "loss": 0.0106, "step": 79390 }, { "grad_norm": 0.14347153902053833, "learning_rate": 1.5429049643606032e-08, "loss": 0.014, "step": 79400 }, { "grad_norm": 0.16373881697654724, "learning_rate": 1.4919900652909935e-08, "loss": 0.0139, "step": 79410 }, { "grad_norm": 0.16574861109256744, "learning_rate": 1.4419292745310175e-08, "loss": 0.0127, "step": 79420 }, { "grad_norm": 0.18988406658172607, "learning_rate": 1.3927226006343885e-08, "loss": 0.0109, "step": 79430 }, { "grad_norm": 0.32573843002319336, "learning_rate": 1.3443700520093805e-08, "loss": 0.0122, "step": 79440 }, { "grad_norm": 0.18049195408821106, "learning_rate": 1.296871636917718e-08, "loss": 0.0081, "step": 79450 }, { "grad_norm": 0.3028578758239746, "learning_rate": 1.2502273634762419e-08, "loss": 0.0115, "step": 79460 }, { "grad_norm": 0.2711257040500641, "learning_rate": 1.2044372396546876e-08, "loss": 0.0122, "step": 79470 }, { "grad_norm": 0.17716418206691742, "learning_rate": 1.1595012732773524e-08, "loss": 0.0132, "step": 79480 }, { "grad_norm": 0.25723016262054443, "learning_rate": 1.1154194720225386e-08, "loss": 0.0136, "step": 79490 }, { "grad_norm": 0.21092121303081512, "learning_rate": 1.0721918434231093e-08, "loss": 0.0112, "step": 79500 }, { "grad_norm": 0.26184728741645813, "learning_rate": 1.0298183948648232e-08, "loss": 0.0124, "step": 79510 }, { "grad_norm": 0.21297359466552734, "learning_rate": 9.88299133588555e-09, "loss": 0.0097, "step": 79520 }, { "grad_norm": 0.19551613926887512, "learning_rate": 9.476340666891847e-09, "loss": 0.015, "step": 79530 }, { "grad_norm": 0.29402339458465576, "learning_rate": 9.078232011139332e-09, "loss": 0.0134, "step": 79540 }, { "grad_norm": 0.28564971685409546, "learning_rate": 8.68866543666802e-09, "loss": 0.0113, "step": 79550 }, { "grad_norm": 0.2981258034706116, "learning_rate": 8.307641010035782e-09, "loss": 0.0132, "step": 79560 }, { "grad_norm": 0.32316961884498596, "learning_rate": 7.93515879635165e-09, "loss": 0.0109, "step": 79570 }, { "grad_norm": 0.6142815947532654, "learning_rate": 7.571218859264706e-09, "loss": 0.0108, "step": 79580 }, { "grad_norm": 0.35524633526802063, "learning_rate": 7.215821260958544e-09, "loss": 0.0119, "step": 79590 }, { "grad_norm": 0.10874711722135544, "learning_rate": 6.868966062162363e-09, "loss": 0.0093, "step": 79600 }, { "grad_norm": 0.10328049957752228, "learning_rate": 6.530653322145419e-09, "loss": 0.0114, "step": 79610 }, { "grad_norm": 0.20506706833839417, "learning_rate": 6.200883098717025e-09, "loss": 0.009, "step": 79620 }, { "grad_norm": 0.321796715259552, "learning_rate": 5.879655448226551e-09, "loss": 0.0157, "step": 79630 }, { "grad_norm": 0.29719802737236023, "learning_rate": 5.566970425557872e-09, "loss": 0.0098, "step": 79640 }, { "grad_norm": 0.19429892301559448, "learning_rate": 5.26282808414047e-09, "loss": 0.0075, "step": 79650 }, { "grad_norm": 0.2908487617969513, "learning_rate": 4.967228475949437e-09, "loss": 0.013, "step": 79660 }, { "grad_norm": 0.26200753450393677, "learning_rate": 4.680171651494369e-09, "loss": 0.0106, "step": 79670 }, { "grad_norm": 0.29527565836906433, "learning_rate": 4.40165765981937e-09, "loss": 0.013, "step": 79680 }, { "grad_norm": 0.13480210304260254, "learning_rate": 4.131686548519698e-09, "loss": 0.0074, "step": 79690 }, { "grad_norm": 0.32300010323524475, "learning_rate": 3.8702583637251214e-09, "loss": 0.0096, "step": 79700 }, { "grad_norm": 0.3600027561187744, "learning_rate": 3.617373150105463e-09, "loss": 0.0152, "step": 79710 }, { "grad_norm": 0.22642338275909424, "learning_rate": 3.3730309508706036e-09, "loss": 0.0114, "step": 79720 }, { "grad_norm": 0.32044556736946106, "learning_rate": 3.137231807781582e-09, "loss": 0.0091, "step": 79730 }, { "grad_norm": 0.24058784544467926, "learning_rate": 2.9099757611172894e-09, "loss": 0.0113, "step": 79740 }, { "grad_norm": 0.2290138155221939, "learning_rate": 2.6912628497133275e-09, "loss": 0.0111, "step": 79750 }, { "grad_norm": 0.23836909234523773, "learning_rate": 2.481093110945354e-09, "loss": 0.0135, "step": 79760 }, { "grad_norm": 0.17304402589797974, "learning_rate": 2.2794665807235327e-09, "loss": 0.014, "step": 79770 }, { "grad_norm": 0.15759356319904327, "learning_rate": 2.0863832934980843e-09, "loss": 0.0113, "step": 79780 }, { "grad_norm": 0.33189892768859863, "learning_rate": 1.901843282264837e-09, "loss": 0.0134, "step": 79790 }, { "grad_norm": 0.12581442296504974, "learning_rate": 1.7258465785541245e-09, "loss": 0.0189, "step": 79800 }, { "grad_norm": 0.15689051151275635, "learning_rate": 1.5583932124418887e-09, "loss": 0.013, "step": 79810 }, { "grad_norm": 0.18248271942138672, "learning_rate": 1.3994832125441282e-09, "loss": 0.009, "step": 79820 }, { "grad_norm": 0.22855417430400848, "learning_rate": 1.2491166060057958e-09, "loss": 0.0118, "step": 79830 }, { "grad_norm": 0.14289115369319916, "learning_rate": 1.1072934185230032e-09, "loss": 0.0123, "step": 79840 }, { "grad_norm": 0.239791139960289, "learning_rate": 9.740136743319195e-10, "loss": 0.0137, "step": 79850 }, { "grad_norm": 0.20241163671016693, "learning_rate": 8.492773962087696e-10, "loss": 0.0089, "step": 79860 }, { "grad_norm": 0.2800082862377167, "learning_rate": 7.330846054587337e-10, "loss": 0.0202, "step": 79870 }, { "grad_norm": 0.3353254199028015, "learning_rate": 6.254353219492526e-10, "loss": 0.0094, "step": 79880 }, { "grad_norm": 0.15502072870731354, "learning_rate": 5.263295640600685e-10, "loss": 0.0107, "step": 79890 }, { "grad_norm": 0.33930474519729614, "learning_rate": 4.357673487387359e-10, "loss": 0.0099, "step": 79900 }, { "grad_norm": 0.20749947428703308, "learning_rate": 3.537486914506616e-10, "loss": 0.0099, "step": 79910 }, { "grad_norm": 0.201485738158226, "learning_rate": 2.8027360621241117e-10, "loss": 0.0128, "step": 79920 }, { "grad_norm": 0.18935856223106384, "learning_rate": 2.153421055806071e-10, "loss": 0.0111, "step": 79930 }, { "grad_norm": 0.14272761344909668, "learning_rate": 1.5895420064637733e-10, "loss": 0.014, "step": 79940 }, { "grad_norm": 0.23987486958503723, "learning_rate": 1.1110990105200891e-10, "loss": 0.0116, "step": 79950 }, { "grad_norm": 0.2527649700641632, "learning_rate": 7.180921496874326e-11, "loss": 0.0147, "step": 79960 }, { "grad_norm": 0.27467861771583557, "learning_rate": 4.10521491134297e-11, "loss": 0.0101, "step": 79970 }, { "grad_norm": 0.3031945526599884, "learning_rate": 1.883870874297422e-11, "loss": 0.0116, "step": 79980 }, { "grad_norm": 0.2541103959083557, "learning_rate": 5.1688976432373584e-12, "loss": 0.0104, "step": 79990 }, { "grad_norm": 0.17103295028209686, "learning_rate": 4.271816234080461e-14, "loss": 0.0078, "step": 80000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }